{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.7" }, "colab": { "name": "10Bar_ANN.ipynb", "provenance": [], "collapsed_sections": [], "toc_visible": true } }, "cells": [ { "cell_type": "markdown", "metadata": { "id": "1MNpyc1WBC41" }, "source": [ "# Data extraction for AI application in Structural Engineering\n", "_Larissa Driemeier, Izabel F. Machado and Gabriel Lopes Rodrigues_\n", " ![](https://drive.google.com/uc?export=view&id=1D5NMNp-KTfou5cSIiDdXwdDDTzRGzToq)\n", "\n", "This introductory notebook replicates the geometry of the structure analysed in the paper\n", "[*Background Information of Deep Learning for Structural Engineering*](https://www.researchgate.net/publication/318190131_Background_Information_of_Deep_Learning_for_Structural_Engineering). \n", "\n", "It is based on the [PMR5251 - Class#3](https://edisciplinas.usp.br/pluginfile.php/5728638/mod_resource/content/1/Aula01_Introd_ML.pdf) and [PMR5251 - Class#4](https://edisciplinas.usp.br/pluginfile.php/5759070/mod_resource/content/1/Aula02_RedesNeuraisArtificiais.pdf)\n" ] }, { "cell_type": "markdown", "metadata": { "id": "bS02BkT-5T63" }, "source": [ "## Geometry\n", "\n", "The figure below shows a beam like 2D truss with 10 bars. The length of the bars are fixed, however the cross sectional areas are obtained through a random uniform sampling between $0.6$ $cm^2$ and $225.8$ $cm^2$. In total, 500 different structures were generated.\n", "\n", " ![](https://drive.google.com/uc?export=view&id=1xOuJYBWiWGkq5l_Z_hcAjYak_hjG26l5)\n", "\n", "Then, the structure is loaded and analysed in the commercial FE software Abaqus. Since the dimensions in the structure are fixed, the input the set of areas, while all nodal displacements and also bar stresses are computed as output." ] }, { "cell_type": "markdown", "metadata": { "id": "57lFpyyBoFtD" }, "source": [ "## Linear material model \n", "\n", "The material characteristics are generic values for Aluminum alloy 6061, as listed below.\n", "\n", "Property | Value                        | Unity                      \n", "--- | --- | ---\n", "Mass density $\\rho$ | $2.768\\times 10^{-9}$ | $ton/mm^2$\n", "Poisson $\\nu$ | $0.35$ | -\n", "Young's Modulus $E$| $68950$ | $MPa$\n", "Yield stress $\\sigma_{y0}$| $200$ | $MPa$\n", "\n", "The material undergoes elastic deformation until it reaches the elastic limit defined by the yield stress. After the elastic limit, the material exhibits plastic behavior,that is, the material deforms irreversibly and does not return to its original shape and size, even when the load is removed. Initially, only elastic behaviour is considered." ] }, { "cell_type": "markdown", "metadata": { "id": "y-ytuiSdBC43" }, "source": [ "## Libraries\n", "Throughout this notebook the new version 2.0 of Tersorflow was used, with built-in keras support, which has been recently released to the public.\n", "To install it, just follow the instructions in [the official website](https://www.tensorflow.org/install), to guarantee that the right version is installed.\n", "\n", "The rest of the libraries used were simply installed using pip, the default Python tool for installing packages. These include:\n", "\n", "- NumPy: library for dealing with large matrices and also providing optimized functions for these data structures\n", "\n", "- Pandas: used to visualize the data and to work with the dataset.\n", "\n", "- matplotlib: used to generate plots from the models.\n", "\n", "- sklearn (also known as scikit-learn): used because of the many useful functions and utilities it provides for machine learning." ] }, { "cell_type": "code", "metadata": { "id": "sMBfXDXABC43", "outputId": "f1b91f9d-c5ba-46a2-dbfe-505ff97b21c4", "colab": { "base_uri": "https://localhost:8080/", "height": 35 } }, "source": [ "import tensorflow as tf\n", "tf.__version__" ], "execution_count": 1, "outputs": [ { "output_type": "execute_result", "data": { "application/vnd.google.colaboratory.intrinsic+json": { "type": "string" }, "text/plain": [ "'2.3.0'" ] }, "metadata": { "tags": [] }, "execution_count": 1 } ] }, { "cell_type": "code", "metadata": { "id": "u5ATF9uyBC47" }, "source": [ "from tensorflow import keras\n", "import numpy as np\n", "import pandas as pd" ], "execution_count": 2, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "HMEnmW28BC49" }, "source": [ "import matplotlib.pyplot as plt\n", "plt.style.use('fivethirtyeight')\n", "%matplotlib inline" ], "execution_count": 3, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "UHa54RD05ude" }, "source": [ "## Data Loading and Manipulation\n", "\n", "Uploading four files:\n", "1. the dataset containing the areas, `areas.csv`;\n", "2. displacements and reaction force along the time, `FinalResult.csv`;\n", "\n", "If you prefer generate new data, we suggest to use the student version of the software [Abaqus](https://edu.3ds.com/en/software/abaqus-student-edition). The following files are available in the same [link](https://edisciplinas.usp.br/course/view.php?id=82602#section-3):\n", " 1. To generate random areas `gera_areas_10.py`;\n", " 2. Script to run in Abaqus to generate data `10-BarStructure.py`;\n", " 3. Basic geometry to be called by the script mentioned in item 2 `BasicInput.inp`;\n", " 4. Copy the file `extracted_data_DATA_HOUR.csv` as `FinalResult.csv` to upload.\n", "\n", " **Important**\n", "\n", "The script in item 02 automatically generates the bar geometry in Abaqus. If you want to build up a geometry - at least once - with Abaqus, Prof Marcilio Alves kindly prepared a tutorial that can be accessed through the [link](https://www.youtube.com/channel/UCEDn-UheEHKLfOKJKmSKzJw). \n" ] }, { "cell_type": "code", "metadata": { "id": "Or0FZJoDncwC", "outputId": "a820ac2e-58c7-4e41-bf55-d875347388be", "colab": { "resources": { "http://localhost:8080/nbextensions/google.colab/files.js": { "data": "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", "ok": true, "headers": [ [ "content-type", "application/javascript" ] ], "status": 200, "status_text": "" } }, "base_uri": "https://localhost:8080/", "height": 108 } }, "source": [ "from google.colab import files\n", "uploaded = files.upload()" ], "execution_count": 4, "outputs": [ { "output_type": "display_data", "data": { "text/html": [ "\n", " \n", " \n", " Upload widget is only available when the cell has been executed in the\n", " current browser session. Please rerun this cell to enable.\n", " \n", " " ], "text/plain": [ "" ] }, "metadata": { "tags": [] } }, { "output_type": "stream", "text": [ "Saving areas.csv to areas.csv\n", "Saving FinalResult.csv to FinalResult.csv\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "82ahz20vBC5C" }, "source": [ "As shown below, there are 10 different areas, which will be the inputs, and various other measurements, which might be used as outputs of the Neural Network" ] }, { "cell_type": "code", "metadata": { "id": "qJAFdYK8BC5A" }, "source": [ "df = pd.read_csv('FinalResult.csv', index_col=0)" ], "execution_count": 5, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "tpm5RpQdBC5D", "outputId": "cd1f1297-8347-4e78-9efa-2c48ef98becd", "colab": { "base_uri": "https://localhost:8080/", "height": 726 } }, "source": [ "df.dtypes" ], "execution_count": 6, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "area1 float64\n", "area2 float64\n", "area3 float64\n", "area4 float64\n", "area5 float64\n", "area6 float64\n", "area7 float64\n", "area8 float64\n", "area9 float64\n", "area10 float64\n", "d1 float64\n", "d2 float64\n", "d3 float64\n", "d4 float64\n", "d5 float64\n", "d6 float64\n", "d7 float64\n", "d8 float64\n", "s11_1 float64\n", "s11_2 float64\n", "s11_3 float64\n", "s11_4 float64\n", "s11_5 float64\n", "s11_6 float64\n", "s11_7 float64\n", "s11_8 float64\n", "s11_9 float64\n", "s11_10 float64\n", "mises_1 float64\n", "mises_2 float64\n", "mises_3 float64\n", "mises_4 float64\n", "mises_5 float64\n", "mises_6 float64\n", "mises_7 float64\n", "mises_8 float64\n", "mises_9 float64\n", "mises_10 float64\n", "dtype: object" ] }, "metadata": { "tags": [] }, "execution_count": 6 } ] }, { "cell_type": "code", "metadata": { "id": "h2P15QlKBC5F", "outputId": "8c4ae811-561a-40ba-a253-d8ed4a2d433f", "colab": { "base_uri": "https://localhost:8080/", "height": 248 } }, "source": [ "# To show all the columns\n", "pd.set_option('display.max_columns', None)\n", "df.head()" ], "execution_count": 7, "outputs": [ { "output_type": "execute_result", "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
area1area2area3area4area5area6area7area8area9area10d1d2d3d4d5d6d7d8s11_1s11_2s11_3s11_4s11_5s11_6s11_7s11_8s11_9s11_10mises_1mises_2mises_3mises_4mises_5mises_6mises_7mises_8mises_9mises_10
iteration
00.0030860.0191440.0172600.0058040.0112170.0101830.0147340.0178220.0021740.00069810.579137-70.774597-24.355942-71.5503926.575743-16.113703-8.060376-16.891342-122876112.05863760.0030187446.05849860.5-60778972.049584148.0-132877080.033294788.0-35960320.092389552.0122876112.05863760.0030187446.05849860.560778972.049584148.0132877080.033294788.035960320.092389552.0
10.0188810.0098060.0172270.0001070.0100900.0163090.0052120.0213470.0203600.0007499.366341-245.959854-17.619713-251.0329446.852213-21.184267-17.206579-26.058153-3115212.036751352.0018957688.038253380.0-129745584.051668864.0-728894336.033372482.0-54035168.0774317952.03115212.036751352.0018957688.038253380.0129745584.051668864.0728894336.033372482.054035168.0774317952.0
20.0006330.0122530.0212100.0086450.0049380.0095660.0007140.0050520.0099210.01122641.207355-185.375214-96.529587-193.67851337.074791-54.188625-68.113937-65.723526-214267184.086978520.0031161468.062610644.0-513610688.0279561120.0-38947984.0-9012399.0-64523128.022189706.0214267184.086978520.0031161468.062610644.0513610688.0279561120.038947984.09012399.064523128.022189706.0
30.0053090.0052590.0049870.0104100.0065860.0005440.0189220.0125910.0145240.0042479.551858-69.896049-15.171593-72.0827948.560319-36.236866-6.773345-43.978748-63326688.058377396.007476665.016489037.0-51074160.064548776.0-36164440.0140273008.0-104346992.045672644.063326688.058377396.007476665.016489037.051074160.064548776.036164440.0140273008.0104346992.045672644.0
40.0224120.0194260.0027820.0075520.0163080.0160760.0211490.0095660.0187520.0151559.622062-46.439331-8.717741-48.5901377.751640-23.164854-7.650749-22.620508-8045611.0-4104622.7514103846.016218065.0-57690196.058450960.0-24679984.056439460.0-58111388.033765824.08045611.04104622.7514103846.016218065.057690196.058450960.024679984.056439460.058111388.033765824.0
\n", "
" ], "text/plain": [ " area1 area2 area3 area4 area5 area6 \\\n", "iteration \n", "0 0.003086 0.019144 0.017260 0.005804 0.011217 0.010183 \n", "1 0.018881 0.009806 0.017227 0.000107 0.010090 0.016309 \n", "2 0.000633 0.012253 0.021210 0.008645 0.004938 0.009566 \n", "3 0.005309 0.005259 0.004987 0.010410 0.006586 0.000544 \n", "4 0.022412 0.019426 0.002782 0.007552 0.016308 0.016076 \n", "\n", " area7 area8 area9 area10 d1 d2 \\\n", "iteration \n", "0 0.014734 0.017822 0.002174 0.000698 10.579137 -70.774597 \n", "1 0.005212 0.021347 0.020360 0.000749 9.366341 -245.959854 \n", "2 0.000714 0.005052 0.009921 0.011226 41.207355 -185.375214 \n", "3 0.018922 0.012591 0.014524 0.004247 9.551858 -69.896049 \n", "4 0.021149 0.009566 0.018752 0.015155 9.622062 -46.439331 \n", "\n", " d3 d4 d5 d6 d7 d8 \\\n", "iteration \n", "0 -24.355942 -71.550392 6.575743 -16.113703 -8.060376 -16.891342 \n", "1 -17.619713 -251.032944 6.852213 -21.184267 -17.206579 -26.058153 \n", "2 -96.529587 -193.678513 37.074791 -54.188625 -68.113937 -65.723526 \n", "3 -15.171593 -72.082794 8.560319 -36.236866 -6.773345 -43.978748 \n", "4 -8.717741 -48.590137 7.751640 -23.164854 -7.650749 -22.620508 \n", "\n", " s11_1 s11_2 s11_3 s11_4 s11_5 \\\n", "iteration \n", "0 -122876112.0 5863760.00 30187446.0 5849860.5 -60778972.0 \n", "1 -3115212.0 36751352.00 18957688.0 38253380.0 -129745584.0 \n", "2 -214267184.0 86978520.00 31161468.0 62610644.0 -513610688.0 \n", "3 -63326688.0 58377396.00 7476665.0 16489037.0 -51074160.0 \n", "4 -8045611.0 -4104622.75 14103846.0 16218065.0 -57690196.0 \n", "\n", " s11_6 s11_7 s11_8 s11_9 s11_10 \\\n", "iteration \n", "0 49584148.0 -132877080.0 33294788.0 -35960320.0 92389552.0 \n", "1 51668864.0 -728894336.0 33372482.0 -54035168.0 774317952.0 \n", "2 279561120.0 -38947984.0 -9012399.0 -64523128.0 22189706.0 \n", "3 64548776.0 -36164440.0 140273008.0 -104346992.0 45672644.0 \n", "4 58450960.0 -24679984.0 56439460.0 -58111388.0 33765824.0 \n", "\n", " mises_1 mises_2 mises_3 mises_4 mises_5 \\\n", "iteration \n", "0 122876112.0 5863760.00 30187446.0 5849860.5 60778972.0 \n", "1 3115212.0 36751352.00 18957688.0 38253380.0 129745584.0 \n", "2 214267184.0 86978520.00 31161468.0 62610644.0 513610688.0 \n", "3 63326688.0 58377396.00 7476665.0 16489037.0 51074160.0 \n", "4 8045611.0 4104622.75 14103846.0 16218065.0 57690196.0 \n", "\n", " mises_6 mises_7 mises_8 mises_9 mises_10 \n", "iteration \n", "0 49584148.0 132877080.0 33294788.0 35960320.0 92389552.0 \n", "1 51668864.0 728894336.0 33372482.0 54035168.0 774317952.0 \n", "2 279561120.0 38947984.0 9012399.0 64523128.0 22189706.0 \n", "3 64548776.0 36164440.0 140273008.0 104346992.0 45672644.0 \n", "4 58450960.0 24679984.0 56439460.0 58111388.0 33765824.0 " ] }, "metadata": { "tags": [] }, "execution_count": 7 } ] }, { "cell_type": "markdown", "metadata": { "id": "-WGv5XThBC5H" }, "source": [ "### Splitting dataset\n", "\n", "The whole dataset will be split into training and test sets. The training set will be used to train the model and the test set to verify its performance.\n", "\n", "\n", "\n" ] }, { "cell_type": "code", "metadata": { "id": "uyOEcTmLBC5I" }, "source": [ "from sklearn.model_selection import train_test_split\n", "train, test = train_test_split(df, test_size=0.2, random_state=1)" ], "execution_count": 8, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "jardg69qBC5K", "outputId": "b065397d-5258-43bc-d071-d7e7e4217e9e", "colab": { "base_uri": "https://localhost:8080/", "height": 248 } }, "source": [ "train.head()" ], "execution_count": 9, "outputs": [ { "output_type": "execute_result", "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
area1area2area3area4area5area6area7area8area9area10d1d2d3d4d5d6d7d8s11_1s11_2s11_3s11_4s11_5s11_6s11_7s11_8s11_9s11_10mises_1mises_2mises_3mises_4mises_5mises_6mises_7mises_8mises_9mises_10
iteration
2010.0209590.0052350.0191970.0100050.0200770.0023490.0012710.0106060.0210140.01053916.004036-120.325661-53.915527-120.18594416.137526-32.663422-50.967178-54.107288-22231914.01.616967e+08-1.006564e+06-1.053551e+06-384316160.0121684416.02838388.5011838953.0-62306460.065863956.022231914.01.616967e+081.006564e+061.053551e+06384316160.0121684416.02838388.5011838953.062306460.065863956.0
920.0225290.0058680.0116310.0167140.0156290.0098220.0175580.0110000.0161720.01112612.547656-53.592022-8.100519-54.67098211.504974-27.936739-6.230652-30.526878-14099667.01.953085e+077.862298e+068.135870e+06-46982004.086752840.0-16162369.0091602400.0-61951560.026877044.014099667.01.953085e+077.862298e+068.135870e+0646982004.086752840.016162369.0091602400.061951560.026877044.0
3440.0032190.0123110.0178840.0161570.0032170.0211460.0210480.0081380.0220800.00435014.224157-54.255360-18.496674-60.12879613.368400-21.245007-6.041500-21.704147-93917784.03.462124e+066.452812e+064.428844e+07-45555708.0100803928.0-46319392.0059051812.0-29696636.026462190.093917784.03.462124e+066.452812e+064.428844e+0745555708.0100803928.046319392.0059051812.029696636.026462190.0
1190.0152170.0201340.0039380.0145350.0110370.0077390.0160590.0220210.0005480.02026718.363760-50.452175-12.587241-51.1659433.985027-24.647602-9.228423-21.760574-25327046.0-2.176952e+071.084223e+085.382156e+06-69586584.030048950.0-4145055.7547249120.0-77902696.037499008.025327046.02.176952e+071.084223e+085.382156e+0669586584.030048950.04145055.7547249120.077902696.037499008.0
2210.0004900.0029860.0065510.0106980.0006630.0015720.0179950.0221350.0097660.0106408.433659-77.570465-70.816208-118.8271415.633955-34.009995-6.182016-55.686005-487371776.01.634472e+082.111106e+073.110945e+08-46615268.042482636.0-27404998.00186641504.0-106984240.031545470.0487371776.01.634472e+082.111106e+073.110945e+0846615268.042482636.027404998.00186641504.0106984240.031545470.0
\n", "
" ], "text/plain": [ " area1 area2 area3 area4 area5 area6 \\\n", "iteration \n", "201 0.020959 0.005235 0.019197 0.010005 0.020077 0.002349 \n", "92 0.022529 0.005868 0.011631 0.016714 0.015629 0.009822 \n", "344 0.003219 0.012311 0.017884 0.016157 0.003217 0.021146 \n", "119 0.015217 0.020134 0.003938 0.014535 0.011037 0.007739 \n", "221 0.000490 0.002986 0.006551 0.010698 0.000663 0.001572 \n", "\n", " area7 area8 area9 area10 d1 d2 \\\n", "iteration \n", "201 0.001271 0.010606 0.021014 0.010539 16.004036 -120.325661 \n", "92 0.017558 0.011000 0.016172 0.011126 12.547656 -53.592022 \n", "344 0.021048 0.008138 0.022080 0.004350 14.224157 -54.255360 \n", "119 0.016059 0.022021 0.000548 0.020267 18.363760 -50.452175 \n", "221 0.017995 0.022135 0.009766 0.010640 8.433659 -77.570465 \n", "\n", " d3 d4 d5 d6 d7 d8 \\\n", "iteration \n", "201 -53.915527 -120.185944 16.137526 -32.663422 -50.967178 -54.107288 \n", "92 -8.100519 -54.670982 11.504974 -27.936739 -6.230652 -30.526878 \n", "344 -18.496674 -60.128796 13.368400 -21.245007 -6.041500 -21.704147 \n", "119 -12.587241 -51.165943 3.985027 -24.647602 -9.228423 -21.760574 \n", "221 -70.816208 -118.827141 5.633955 -34.009995 -6.182016 -55.686005 \n", "\n", " s11_1 s11_2 s11_3 s11_4 s11_5 \\\n", "iteration \n", "201 -22231914.0 1.616967e+08 -1.006564e+06 -1.053551e+06 -384316160.0 \n", "92 -14099667.0 1.953085e+07 7.862298e+06 8.135870e+06 -46982004.0 \n", "344 -93917784.0 3.462124e+06 6.452812e+06 4.428844e+07 -45555708.0 \n", "119 -25327046.0 -2.176952e+07 1.084223e+08 5.382156e+06 -69586584.0 \n", "221 -487371776.0 1.634472e+08 2.111106e+07 3.110945e+08 -46615268.0 \n", "\n", " s11_6 s11_7 s11_8 s11_9 s11_10 \\\n", "iteration \n", "201 121684416.0 2838388.50 11838953.0 -62306460.0 65863956.0 \n", "92 86752840.0 -16162369.00 91602400.0 -61951560.0 26877044.0 \n", "344 100803928.0 -46319392.00 59051812.0 -29696636.0 26462190.0 \n", "119 30048950.0 -4145055.75 47249120.0 -77902696.0 37499008.0 \n", "221 42482636.0 -27404998.00 186641504.0 -106984240.0 31545470.0 \n", "\n", " mises_1 mises_2 mises_3 mises_4 mises_5 \\\n", "iteration \n", "201 22231914.0 1.616967e+08 1.006564e+06 1.053551e+06 384316160.0 \n", "92 14099667.0 1.953085e+07 7.862298e+06 8.135870e+06 46982004.0 \n", "344 93917784.0 3.462124e+06 6.452812e+06 4.428844e+07 45555708.0 \n", "119 25327046.0 2.176952e+07 1.084223e+08 5.382156e+06 69586584.0 \n", "221 487371776.0 1.634472e+08 2.111106e+07 3.110945e+08 46615268.0 \n", "\n", " mises_6 mises_7 mises_8 mises_9 mises_10 \n", "iteration \n", "201 121684416.0 2838388.50 11838953.0 62306460.0 65863956.0 \n", "92 86752840.0 16162369.00 91602400.0 61951560.0 26877044.0 \n", "344 100803928.0 46319392.00 59051812.0 29696636.0 26462190.0 \n", "119 30048950.0 4145055.75 47249120.0 77902696.0 37499008.0 \n", "221 42482636.0 27404998.00 186641504.0 106984240.0 31545470.0 " ] }, "metadata": { "tags": [] }, "execution_count": 9 } ] }, { "cell_type": "code", "metadata": { "id": "acRAZpUHBC5M", "outputId": "f65c2901-fcc8-4769-f342-cf5632e64c41", "colab": { "base_uri": "https://localhost:8080/", "height": 35 } }, "source": [ "train.shape, test.shape" ], "execution_count": 10, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "((416, 38), (104, 38))" ] }, "metadata": { "tags": [] }, "execution_count": 10 } ] }, { "cell_type": "markdown", "metadata": { "id": "AtoamKYSBC5O" }, "source": [ "### Defining the training values and the expected outputs\n", "\n", "The input for the NN is a vector with all 10 areas that compound the geometry of the structure. Remember that we mantain all other parameters, such as material properties and dimensions, fixed.\n", "\n", "As output for the NN, let's generate an array of displacements (\\[*d2*, *d4*\\]). These correspond to the vertical displacements of the rightmost nodes of the structure.\n" ] }, { "cell_type": "code", "metadata": { "id": "2l5vD_siBC5P" }, "source": [ "x_train = train.loc[:,'area1':'area10'].values\n", "y_train = train[['d2', 'd4']].values\n", "x_test = test.loc[:,'area1':'area10'].values\n", "y_test = test[['d2', 'd4']].values" ], "execution_count": 11, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "5G27Ww8pvUC-" }, "source": [ "### FEA results\n", "\n", "Read the results from FEA, where `d2` and `d4` are the displacements at the rightmost nodes of the structure to see the variation our future NN has to learn." ] }, { "cell_type": "code", "metadata": { "id": "9JW7edpTvEjK", "outputId": "7ef3e60a-21bf-4e06-fcd9-a9cf7822c0e2", "colab": { "base_uri": "https://localhost:8080/", "height": 279 } }, "source": [ "disp2 = df[['d2']].values\n", "disp4 = df[['d4']].values\n", "xAxis = [i + 1.0 for i, _ in enumerate(disp2)]\n", "plt.scatter(xAxis,disp2,color='firebrick',s=6, label =r'$d_2$')\n", "plt.scatter(xAxis,disp4,color='darkslateblue',s=8, label =r'$d_4$')\n", "plt.title('Displacements at the end of the truss structure')\n", "plt.legend()\n", "plt.show()" ], "execution_count": 12, "outputs": [ { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [], "needs_background": "light" } } ] }, { "cell_type": "markdown", "metadata": { "id": "NZHeVHAdBC5R" }, "source": [ "### Normalizing Dataset\n", "\n", "As the paper has done, the input data was normalized, to obtain better results.\n", "\n", "The normalizer of the sklearn library was used to facilitate the process." ] }, { "cell_type": "code", "metadata": { "id": "CRF7EoS0BC5S" }, "source": [ "from sklearn import preprocessing\n", "\n", "normalizer = preprocessing.StandardScaler().fit(x_train)\n", "x_train_norm = normalizer.transform(x_train)\n", "x_test_norm = normalizer.transform(x_test)" ], "execution_count": 13, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "uEQDFdR0BC5U" }, "source": [ "As can be seen below, the mean and standard deviation for the training set are, respectively, approximately 0 and 1. Since the dataset is normalized with respect to the training set, \n", "for the test data the mean and standard deviation are, in this case, slightly different." ] }, { "cell_type": "code", "metadata": { "id": "GQjQ8vqRBC5U", "outputId": "686fbf6f-f17a-4130-fb2a-a5dd0259fbbe", "colab": { "base_uri": "https://localhost:8080/", "height": 54 } }, "source": [ "# Mean\n", "mean_train = np.mean(x_train_norm)\n", "mean_test = np.mean(x_test_norm)\n", "\n", "# Standard Deviation\n", "std_train = np.std(x_train_norm)\n", "std_test = np.std(x_test_norm)\n", "\n", "print('For the training set, the mean is {:2.4f} and the standard deviation is {:2.4f}'.format(mean_train,std_train))\n", "print('For the test set, the mean is {:2.4f} and the standard deviation is {:2.4f}'.format(mean_test,std_test))\n" ], "execution_count": 14, "outputs": [ { "output_type": "stream", "text": [ "For the training set, the mean is -0.0000 and the standard deviation is 1.0000\n", "For the test set, the mean is -0.0043 and the standard deviation is 0.9993\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "QzKWzeY1iNQ9" }, "source": [ "# Training and evaluation" ] }, { "cell_type": "markdown", "metadata": { "id": "DlNNpHuqiQ6Q" }, "source": [ "## First model\n", "\n", "This model is a Neural Network with architecture (10-20-2), sigmoid activation function and Stochastic Gradient Descendent (SGD) as the optimizer. It is exactly the first model described in the paper.\n", "\n" ] }, { "cell_type": "markdown", "metadata": { "id": "P1h4_elMwSxK" }, "source": [ "Building deep-learning models in Keras is\n", "done by clipping together compatible layers to form useful data-transformation pipelines.\n", "The notion of layer compatibility here refers specifically to the fact that every layer\n", "will only accept input tensors of a certain shape and will return output tensors of a certain\n", "shape.\n", "\n", "Therefore, the architecture (10-20-2) our network consists of one input layer with 10 features, followed by a sequence of two Dense layers, which are fully connected neural layers. The first hidden layer has 20 neurons, and the second (and last) layer has 2 neurons." ] }, { "cell_type": "code", "metadata": { "id": "npWzQbmUisbm" }, "source": [ "from keras import models\n", "from keras import layers\n", "\n", "model = models.Sequential()\n", "model.add(layers.Dense(20, activation='sigmoid', input_shape=(10,)))\n", "model.add(layers.Dense(2))" ], "execution_count": 15, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "PkGkuWT-wkaB" }, "source": [ "To make the network ready for training, we need to pick three more things, as part\n", "of the compilation step:\n", "+ The loss function `mean_squared_error` — How the network will be able to measure its performance on the training data, and thus how it will be able to steer itself in the right direction. We choose \n", "+ The optimizer `sgd` — The mechanism through which the network will update itself based on the data it sees and its loss function.\n", "+ Metrics to monitor during training and testing `mean_absolute_error`, `mean_absolute_percentage_error`.\n", "\n", "### Loss Function\n", "\n", "Mean squared error $E$ is calculated as the average of the squared differences between the predicted $\\hat{\\mathbf{y}}^{(i)}$ and target values ${\\mathbf{y}}^{(i)}$,\n", "$$\n", "E(\\hat{\\mathbf{y}}^{(i)},{\\mathbf{y}}^{(i)})=\\sum\\limits_{j=1}^{n_y} \\left(\\hat{y}^{(i)}_j -{y}^{(i)}_j\\right)^2= \\left\\|\\hat{\\mathbf{y}}^{(i)} - \\hat{\\mathbf{y}}^{(i)} \\right\\|_2^2\n", "$$\n", "for the data $i$, $i=1,...m$.\n", "Then, the loss function $J$\n", "$$\n", "J\\left(\\mathbf{W},\\mathbf{B}\\right)={1\\over m}\\sum\\limits_{i=1}^{m}E(\\hat{\\mathbf{y}}^{(i)},{\\mathbf{y}}^{(i)}) = {1\\over m}\\sum\\limits_{i=1}^{m}\\sum\\limits_{j=1}^{n_y} \\left(\\hat{y}^{(i)}_j -{y}^{(i)}_j\\right)^2 = {1\\over m} \\sum\\limits_{i=1}^{m}\\left\\|\\hat{\\mathbf{y}}^{(i)} - \\hat{\\mathbf{y}}^{(i)} \\right\\|_2^2\n", "$$\n", "depends on the weights $\\mathbf{W}$ and bias $\\mathbf{b}$ parameters.\n", "The result is always positive regardless of the sign of the predicted and actual values and a perfect value is 0.0. \n", "\n", "The squaring means that larger mistakes result in more error than smaller mistakes, that is, the model is punished for making larger mistakes.\n", "\n", "The mean squared error loss function can be used in Keras by specifying ‘mse‘ or `mean_squared_error` as the loss function when compiling the model.\n", "\n", "### Optimizer\n", "\n", "SGD is the same as gradient descent, except that it is used for only partial data to train every time. The parameter is called *mini-batch size*.\n", "\n", "Faster optimizers are available in the literature to speed up the training step. We will apply the SGD + Momentum (known as SGD), but, be aware that are other popular Optimizer approaches aush as Nesterov Accelerated Gradient, AdaGrad, RMSProp, Adam, and Nadam optimization.\n", "\n", "The two recommended updates to use are either SGD+Nesterov Momentum or Adam.\n", "\n", "The SGD optimizer used in the paper had a learning rate of 0.01 and momentum of 0.9, which probably due to differences in the dataset, didn't converge. To achieve better results, a learning rate of 0.001 was used instead.\n", "\n", "### Metrics\n", "\n", "A metric or Key Performance Indicator (KPI) is a function that is used to judge the performance of your model. The most commonly used are defined below.\n", "\n", "**MAE**\n", "\n", "The Mean Absolute Error (`mean_absolute_error`,`MAE`, `mae`) computes the mean absolute error between the labels and predictions \n", "$$\n", "MAE = \\frac{1}{n} \\sum_1^n |y^{(i)} - \\hat{y}^{(i)}|\n", "$$\n", "\n", "\n", "**MAPE**\n", "\n", "The Mean Absolute Percentage Error (`mean_absolute_percentage_error`, `MAPE`, `mape`) is one of which is\n", "$$\n", "MAPE = \\frac{100}{n} \\sum_i^n \\frac{y^{(i)} - \\hat{y}^{(i)}}{y^{(i)}}\n", "$$\n", "\n", "Similar to MAE, but normalized by true observation. Downside is when true observation value $\\hat{y}^{(i)}$ is zero or near to zero, this metric will be problematic.\n", "\n", "**MSE**\n", "\n", "Mean squared error (`mean_squared_error`, `MSE` or `mse`) is a quadratic scoring rule that also measures the average magnitude of the error. It’s the average of squared differences between prediction and actual observation,\n", "$$\n", "MSE =\\frac{1}{n} \\sum_i^n (y^{(i)} -\\hat{y}^{(i)})^2\n", "$$\n", "\n", "MSE is like a combination measurement of bias and variance of your prediction, i.e., $MSE = Bias^2 + Var$.\n", "\n", "MAE and MSE are two of the most common metrics used to measure accuracy for continuous variables. They express average model prediction error in units of the variable of interest, can range from $0$ to $+\\infty$ and are indifferent to the direction of errors. They are negatively-oriented scores, which means lower values are better.\n", "\n", "Taking the average of the squared errors has some interesting implications for MSE. Since the errors are squared, the RMSE gives a relatively high weight to large errors. This means the MSE should be more useful when large errors are particularly undesirable. \n" ] }, { "cell_type": "code", "metadata": { "id": "NwlBi4P-wnep" }, "source": [ "from tensorflow.keras import optimizers\n", "\n", "sgd = optimizers.SGD(lr=0.001, momentum=0.9)\n", "\n", "model.compile(optimizer=sgd,\n", " loss='mean_squared_error',\n", " metrics=['mean_absolute_error', 'mean_absolute_percentage_error'])" ], "execution_count": 16, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "B4dj_21ji8DI" }, "source": [ "The code below is a callback to create a loss history for the test set. It evaluates the model after each epoch and appends the result into an array. It is used to generate a plot of the model loss for both the training and the test sets. \n", "\n", "However, it is also very slow and makes training last for more than half an hour with large epochs, which is why the code to call it is normally commented." ] }, { "cell_type": "code", "metadata": { "id": "fKATE-Cti6r_" }, "source": [ "class TestLossHistory(keras.callbacks.Callback):\n", " def __init__(self, x_test, y_test):\n", " self.x_test = x_test\n", " self.y_test = y_test\n", " self.i = 0\n", " def on_train_begin(self, logs={}):\n", " self.losses = []\n", " def on_epoch_end(self, batch, logs={}):\n", " #print(f\"logs: {logs}\")\n", " self.losses.append(self.model.evaluate(self.x_test, self.y_test))\n", " def on_train_end(self, logs={}):\n", " self.losses = np.array(self.losses)" ], "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "E8-AUpRmkAQ-" }, "source": [ "\n", "### Gradient descent\n", "\n", "The Gradient Descent method is the basic *motor* of artificial neural networks. \n", "As discussed in class, it is an iterative method and, therefore, training a deep neural network can be an extremely time-consuming task especially with complex problems. \n", "\n", "There are other versions of Gradient Descent. As far as the size of the dataset used to update parameters is concern, the options are,\n", "* Batch Gradient Descent (BGD, or simply GD) - running on a full dataset. Gradient is more general, but intractable for huge datasets;\n", "* Stochastic Gradient Descent (SGD) - picking a random instance at each step. Gradient can be noisy;\n", "* Mini-batch Gradient Descent (MBGD) - running on random subsets of a dataset - looking for a balance between BGD and SGD. Not very noisy and computationally tractable too, that means, best of both worlds.\n", "\n", "As usually, there is no free lunch and all have pros and cons. Batch Gradient Descent can reach the global minimum at a terribly slow pace, specially with huge problems since in modern day architectures, the number of parameters may be in billions. Mini-batch Gradient Descent gets to the global minimum faster than BGD but it is easier to get stuck in the local minimum, and SGD is usually harder to get to the global minimum compared to the other two.\n", "\n", "We'll train our NN using MBGD and BGD, in order to compare effectiveness.\n", "\n", "In Keras `batch_size` refers to the batch size in MBGD. If you want to run a BGD, you need to set the `batch_size` to the number of training samples.\n", "\n", "Below, train and test are performed with mini-batch. Teste com 10000 epochs." ] }, { "cell_type": "code", "metadata": { "id": "VyHRZYqrmnwc", "outputId": "54db4e7f-ba90-48f4-e751-d7aec751e857", "colab": { "base_uri": "https://localhost:8080/", "height": 1000 } }, "source": [ "history_with_minibatch = model.fit(x_train_norm, y_train, epochs=1000, batch_size=32)\n", "\n", "# To use the test loss history, comment the lines above and uncomment the lines below\n", "#test_history_with_minibatch = TestLossHistory(x_test, y_test)\n", "#history_with_minibatch = model.fit(x_train, y_train, epochs=10000, batch_size=32, \n", "# callbacks=[test_history_with_minibatch])\n", "\n" ], "execution_count": 17, "outputs": [ { "output_type": "stream", "text": [ "Epoch 1/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 10621.5059 - mean_absolute_error: 82.8233 - mean_absolute_percentage_error: 82.0616\n", "Epoch 2/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 3606.0720 - mean_absolute_error: 39.8355 - mean_absolute_percentage_error: 39.4021\n", "Epoch 3/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 2801.1323 - mean_absolute_error: 39.5541 - mean_absolute_percentage_error: 44.2551\n", "Epoch 4/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 2477.4731 - mean_absolute_error: 33.5589 - mean_absolute_percentage_error: 34.7027\n", "Epoch 5/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 2205.7932 - mean_absolute_error: 28.9200 - mean_absolute_percentage_error: 27.3565\n", "Epoch 6/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 2033.7925 - mean_absolute_error: 27.7480 - mean_absolute_percentage_error: 26.0187\n", "Epoch 7/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1898.1995 - mean_absolute_error: 26.3663 - mean_absolute_percentage_error: 24.5462\n", "Epoch 8/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1807.7755 - mean_absolute_error: 24.9997 - mean_absolute_percentage_error: 22.8413\n", "Epoch 9/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1713.3452 - mean_absolute_error: 24.4800 - mean_absolute_percentage_error: 22.4782\n", "Epoch 10/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1663.8147 - mean_absolute_error: 23.9336 - mean_absolute_percentage_error: 21.9247\n", "Epoch 11/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1579.4285 - mean_absolute_error: 23.9583 - mean_absolute_percentage_error: 22.2054\n", "Epoch 12/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1504.1570 - mean_absolute_error: 23.4518 - mean_absolute_percentage_error: 21.8619\n", "Epoch 13/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 1523.0854 - mean_absolute_error: 22.9345 - mean_absolute_percentage_error: 21.0876\n", "Epoch 14/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1459.1057 - mean_absolute_error: 22.9902 - mean_absolute_percentage_error: 21.4573\n", "Epoch 15/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1418.6309 - mean_absolute_error: 22.1725 - mean_absolute_percentage_error: 20.3149\n", "Epoch 16/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1383.3905 - mean_absolute_error: 22.1328 - mean_absolute_percentage_error: 20.7391\n", "Epoch 17/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1343.2765 - mean_absolute_error: 21.7236 - mean_absolute_percentage_error: 20.1505\n", "Epoch 18/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1317.6267 - mean_absolute_error: 21.8978 - mean_absolute_percentage_error: 20.5051\n", "Epoch 19/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1295.6313 - mean_absolute_error: 21.0811 - mean_absolute_percentage_error: 19.3412\n", "Epoch 20/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1291.3068 - mean_absolute_error: 21.3318 - mean_absolute_percentage_error: 19.7496\n", "Epoch 21/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1244.2189 - mean_absolute_error: 21.0227 - mean_absolute_percentage_error: 19.6265\n", "Epoch 22/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1201.2894 - mean_absolute_error: 20.4551 - mean_absolute_percentage_error: 19.0394\n", "Epoch 23/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1165.1614 - mean_absolute_error: 20.3833 - mean_absolute_percentage_error: 19.1382\n", "Epoch 24/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1158.2805 - mean_absolute_error: 20.1843 - mean_absolute_percentage_error: 18.8217\n", "Epoch 25/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1123.9601 - mean_absolute_error: 20.0790 - mean_absolute_percentage_error: 18.9518\n", "Epoch 26/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1104.7620 - mean_absolute_error: 19.8366 - mean_absolute_percentage_error: 18.6666\n", "Epoch 27/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1079.4095 - mean_absolute_error: 19.5343 - mean_absolute_percentage_error: 18.3194\n", "Epoch 28/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1087.9408 - mean_absolute_error: 19.3867 - mean_absolute_percentage_error: 17.9950\n", "Epoch 29/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1049.3031 - mean_absolute_error: 19.7594 - mean_absolute_percentage_error: 19.0077\n", "Epoch 30/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1042.2898 - mean_absolute_error: 19.0751 - mean_absolute_percentage_error: 17.9517\n", "Epoch 31/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1020.3265 - mean_absolute_error: 19.0686 - mean_absolute_percentage_error: 18.0990\n", "Epoch 32/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1006.3961 - mean_absolute_error: 19.2628 - mean_absolute_percentage_error: 18.4564\n", "Epoch 33/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1003.3375 - mean_absolute_error: 18.6293 - mean_absolute_percentage_error: 17.5011\n", "Epoch 34/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 1014.4600 - mean_absolute_error: 19.4499 - mean_absolute_percentage_error: 18.7413\n", "Epoch 35/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 988.9056 - mean_absolute_error: 18.7127 - mean_absolute_percentage_error: 17.8059\n", "Epoch 36/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 967.3329 - mean_absolute_error: 18.6799 - mean_absolute_percentage_error: 17.7362\n", "Epoch 37/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 953.6013 - mean_absolute_error: 19.0199 - mean_absolute_percentage_error: 18.4808\n", "Epoch 38/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 949.6829 - mean_absolute_error: 18.6215 - mean_absolute_percentage_error: 17.8608\n", "Epoch 39/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 955.7107 - mean_absolute_error: 18.2415 - mean_absolute_percentage_error: 17.0985\n", "Epoch 40/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 932.9346 - mean_absolute_error: 18.7786 - mean_absolute_percentage_error: 18.2217\n", "Epoch 41/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 907.8183 - mean_absolute_error: 18.2451 - mean_absolute_percentage_error: 17.4734\n", "Epoch 42/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 938.2470 - mean_absolute_error: 18.1055 - mean_absolute_percentage_error: 17.1165\n", "Epoch 43/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 907.0062 - mean_absolute_error: 18.2864 - mean_absolute_percentage_error: 17.5848\n", "Epoch 44/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 894.2353 - mean_absolute_error: 17.8418 - mean_absolute_percentage_error: 16.8265\n", "Epoch 45/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 886.0704 - mean_absolute_error: 17.9497 - mean_absolute_percentage_error: 17.2564\n", "Epoch 46/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 869.5094 - mean_absolute_error: 17.7187 - mean_absolute_percentage_error: 17.0011\n", "Epoch 47/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 871.8974 - mean_absolute_error: 17.6166 - mean_absolute_percentage_error: 16.6153\n", "Epoch 48/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 867.9619 - mean_absolute_error: 17.8787 - mean_absolute_percentage_error: 17.2049\n", "Epoch 49/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 855.1356 - mean_absolute_error: 17.4217 - mean_absolute_percentage_error: 16.4823\n", "Epoch 50/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 843.5758 - mean_absolute_error: 17.4971 - mean_absolute_percentage_error: 16.7355\n", "Epoch 51/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 837.1049 - mean_absolute_error: 17.5768 - mean_absolute_percentage_error: 17.0772\n", "Epoch 52/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 828.7719 - mean_absolute_error: 17.1010 - mean_absolute_percentage_error: 16.3021\n", "Epoch 53/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 817.1689 - mean_absolute_error: 17.2214 - mean_absolute_percentage_error: 16.6241\n", "Epoch 54/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 832.1257 - mean_absolute_error: 17.0355 - mean_absolute_percentage_error: 16.1493\n", "Epoch 55/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 818.6218 - mean_absolute_error: 17.1240 - mean_absolute_percentage_error: 16.4292\n", "Epoch 56/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 810.8812 - mean_absolute_error: 17.2246 - mean_absolute_percentage_error: 16.5501\n", "Epoch 57/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 793.1439 - mean_absolute_error: 16.9688 - mean_absolute_percentage_error: 16.4368\n", "Epoch 58/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 790.9196 - mean_absolute_error: 16.7083 - mean_absolute_percentage_error: 16.0709\n", "Epoch 59/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 792.3934 - mean_absolute_error: 16.8877 - mean_absolute_percentage_error: 16.2824\n", "Epoch 60/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 772.5443 - mean_absolute_error: 16.5304 - mean_absolute_percentage_error: 15.7858\n", "Epoch 61/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 795.5968 - mean_absolute_error: 16.8120 - mean_absolute_percentage_error: 16.3203\n", "Epoch 62/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 763.5688 - mean_absolute_error: 16.5732 - mean_absolute_percentage_error: 16.0292\n", "Epoch 63/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 756.9969 - mean_absolute_error: 16.3107 - mean_absolute_percentage_error: 15.7248\n", "Epoch 64/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 750.0146 - mean_absolute_error: 16.1945 - mean_absolute_percentage_error: 15.6056\n", "Epoch 65/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 744.5909 - mean_absolute_error: 16.2607 - mean_absolute_percentage_error: 15.7019\n", "Epoch 66/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 735.7865 - mean_absolute_error: 16.0583 - mean_absolute_percentage_error: 15.4840\n", "Epoch 67/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 730.1110 - mean_absolute_error: 16.0906 - mean_absolute_percentage_error: 15.6554\n", "Epoch 68/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 721.2408 - mean_absolute_error: 15.9877 - mean_absolute_percentage_error: 15.5826\n", "Epoch 69/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 718.8966 - mean_absolute_error: 15.9354 - mean_absolute_percentage_error: 15.5110\n", "Epoch 70/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 726.6718 - mean_absolute_error: 15.7488 - mean_absolute_percentage_error: 15.1441\n", "Epoch 71/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 713.8696 - mean_absolute_error: 15.9434 - mean_absolute_percentage_error: 15.5813\n", "Epoch 72/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 714.5927 - mean_absolute_error: 15.8699 - mean_absolute_percentage_error: 15.4620\n", "Epoch 73/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 704.5319 - mean_absolute_error: 15.7957 - mean_absolute_percentage_error: 15.3830\n", "Epoch 74/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 706.9695 - mean_absolute_error: 15.8842 - mean_absolute_percentage_error: 15.5747\n", "Epoch 75/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 696.8457 - mean_absolute_error: 15.6999 - mean_absolute_percentage_error: 15.3268\n", "Epoch 76/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 692.8262 - mean_absolute_error: 15.6130 - mean_absolute_percentage_error: 15.2454\n", "Epoch 77/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 690.9757 - mean_absolute_error: 15.5532 - mean_absolute_percentage_error: 15.1605\n", "Epoch 78/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 685.8174 - mean_absolute_error: 15.7929 - mean_absolute_percentage_error: 15.6704\n", "Epoch 79/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 680.7406 - mean_absolute_error: 15.6912 - mean_absolute_percentage_error: 15.5125\n", "Epoch 80/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 694.1395 - mean_absolute_error: 15.5510 - mean_absolute_percentage_error: 15.1679\n", "Epoch 81/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 686.0966 - mean_absolute_error: 15.7736 - mean_absolute_percentage_error: 15.5940\n", "Epoch 82/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 665.8988 - mean_absolute_error: 15.4705 - mean_absolute_percentage_error: 15.2964\n", "Epoch 83/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 674.9686 - mean_absolute_error: 15.4214 - mean_absolute_percentage_error: 15.1959\n", "Epoch 84/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 668.2223 - mean_absolute_error: 15.5630 - mean_absolute_percentage_error: 15.4432\n", "Epoch 85/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 672.4905 - mean_absolute_error: 15.4928 - mean_absolute_percentage_error: 15.2761\n", "Epoch 86/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 667.1877 - mean_absolute_error: 15.3232 - mean_absolute_percentage_error: 15.0689\n", "Epoch 87/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 663.1334 - mean_absolute_error: 15.4436 - mean_absolute_percentage_error: 15.2412\n", "Epoch 88/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 655.6500 - mean_absolute_error: 15.4188 - mean_absolute_percentage_error: 15.3645\n", "Epoch 89/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 667.6923 - mean_absolute_error: 15.3211 - mean_absolute_percentage_error: 15.1052\n", "Epoch 90/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 664.7216 - mean_absolute_error: 15.5005 - mean_absolute_percentage_error: 15.4287\n", "Epoch 91/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 645.5712 - mean_absolute_error: 15.2061 - mean_absolute_percentage_error: 15.1331\n", "Epoch 92/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 657.6842 - mean_absolute_error: 15.0526 - mean_absolute_percentage_error: 14.7714\n", "Epoch 93/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 639.9222 - mean_absolute_error: 15.2592 - mean_absolute_percentage_error: 15.2650\n", "Epoch 94/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 641.6832 - mean_absolute_error: 15.3916 - mean_absolute_percentage_error: 15.5000\n", "Epoch 95/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 647.5839 - mean_absolute_error: 15.1354 - mean_absolute_percentage_error: 14.9576\n", "Epoch 96/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 645.7848 - mean_absolute_error: 15.2172 - mean_absolute_percentage_error: 15.1313\n", "Epoch 97/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 642.9482 - mean_absolute_error: 15.2477 - mean_absolute_percentage_error: 15.1563\n", "Epoch 98/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 648.5294 - mean_absolute_error: 15.2705 - mean_absolute_percentage_error: 15.2542\n", "Epoch 99/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 632.3431 - mean_absolute_error: 15.1373 - mean_absolute_percentage_error: 15.1155\n", "Epoch 100/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 626.5630 - mean_absolute_error: 15.1270 - mean_absolute_percentage_error: 15.1115\n", "Epoch 101/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 628.8591 - mean_absolute_error: 15.0615 - mean_absolute_percentage_error: 15.0106\n", "Epoch 102/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 623.9133 - mean_absolute_error: 15.0593 - mean_absolute_percentage_error: 15.1032\n", "Epoch 103/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 610.3461 - mean_absolute_error: 15.0124 - mean_absolute_percentage_error: 15.0758\n", "Epoch 104/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 616.7584 - mean_absolute_error: 14.9287 - mean_absolute_percentage_error: 14.8423\n", "Epoch 105/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 606.9988 - mean_absolute_error: 14.8657 - mean_absolute_percentage_error: 14.8334\n", "Epoch 106/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 603.3060 - mean_absolute_error: 14.9684 - mean_absolute_percentage_error: 15.1182\n", "Epoch 107/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 607.0275 - mean_absolute_error: 14.8384 - mean_absolute_percentage_error: 14.8885\n", "Epoch 108/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 604.9166 - mean_absolute_error: 14.9800 - mean_absolute_percentage_error: 15.0593\n", "Epoch 109/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 599.9544 - mean_absolute_error: 14.8170 - mean_absolute_percentage_error: 14.9091\n", "Epoch 110/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 593.4267 - mean_absolute_error: 14.7372 - mean_absolute_percentage_error: 14.8136\n", "Epoch 111/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 589.8007 - mean_absolute_error: 14.7198 - mean_absolute_percentage_error: 14.8542\n", "Epoch 112/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 588.7153 - mean_absolute_error: 14.7856 - mean_absolute_percentage_error: 14.9522\n", "Epoch 113/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 596.5489 - mean_absolute_error: 14.8024 - mean_absolute_percentage_error: 14.9651\n", "Epoch 114/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 590.1845 - mean_absolute_error: 14.6769 - mean_absolute_percentage_error: 14.7600\n", "Epoch 115/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 583.6869 - mean_absolute_error: 14.6611 - mean_absolute_percentage_error: 14.8379\n", "Epoch 116/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 577.6613 - mean_absolute_error: 14.5966 - mean_absolute_percentage_error: 14.8171\n", "Epoch 117/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 579.4330 - mean_absolute_error: 14.5876 - mean_absolute_percentage_error: 14.7855\n", "Epoch 118/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 575.4009 - mean_absolute_error: 14.6021 - mean_absolute_percentage_error: 14.8648\n", "Epoch 119/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 578.8060 - mean_absolute_error: 14.6123 - mean_absolute_percentage_error: 14.7928\n", "Epoch 120/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 574.1100 - mean_absolute_error: 14.4312 - mean_absolute_percentage_error: 14.6360\n", "Epoch 121/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 569.3292 - mean_absolute_error: 14.4956 - mean_absolute_percentage_error: 14.7450\n", "Epoch 122/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 567.4817 - mean_absolute_error: 14.5137 - mean_absolute_percentage_error: 14.7984\n", "Epoch 123/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 569.0927 - mean_absolute_error: 14.3497 - mean_absolute_percentage_error: 14.5309\n", "Epoch 124/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 569.3541 - mean_absolute_error: 14.5295 - mean_absolute_percentage_error: 14.7659\n", "Epoch 125/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 563.3040 - mean_absolute_error: 14.5667 - mean_absolute_percentage_error: 14.9265\n", "Epoch 126/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 568.5435 - mean_absolute_error: 14.4518 - mean_absolute_percentage_error: 14.7446\n", "Epoch 127/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 565.6354 - mean_absolute_error: 14.4235 - mean_absolute_percentage_error: 14.7021\n", "Epoch 128/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 570.8193 - mean_absolute_error: 14.5579 - mean_absolute_percentage_error: 14.8782\n", "Epoch 129/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 558.6598 - mean_absolute_error: 14.4695 - mean_absolute_percentage_error: 14.8228\n", "Epoch 130/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 558.9679 - mean_absolute_error: 14.4383 - mean_absolute_percentage_error: 14.7779\n", "Epoch 131/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 562.5416 - mean_absolute_error: 14.4132 - mean_absolute_percentage_error: 14.7145\n", "Epoch 132/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 557.7396 - mean_absolute_error: 14.3370 - mean_absolute_percentage_error: 14.6792\n", "Epoch 133/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 553.8339 - mean_absolute_error: 14.3529 - mean_absolute_percentage_error: 14.7929\n", "Epoch 134/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 550.4344 - mean_absolute_error: 14.2608 - mean_absolute_percentage_error: 14.6112\n", "Epoch 135/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 547.8499 - mean_absolute_error: 14.1854 - mean_absolute_percentage_error: 14.4663\n", "Epoch 136/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 547.0052 - mean_absolute_error: 14.3647 - mean_absolute_percentage_error: 14.8161\n", "Epoch 137/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 546.6331 - mean_absolute_error: 14.2450 - mean_absolute_percentage_error: 14.6514\n", "Epoch 138/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 545.8932 - mean_absolute_error: 14.2589 - mean_absolute_percentage_error: 14.6713\n", "Epoch 139/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 552.0403 - mean_absolute_error: 14.1581 - mean_absolute_percentage_error: 14.4303\n", "Epoch 140/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 551.8979 - mean_absolute_error: 14.4467 - mean_absolute_percentage_error: 14.9261\n", "Epoch 141/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 543.5722 - mean_absolute_error: 14.3163 - mean_absolute_percentage_error: 14.7920\n", "Epoch 142/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 547.1868 - mean_absolute_error: 14.2198 - mean_absolute_percentage_error: 14.7074\n", "Epoch 143/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 537.9515 - mean_absolute_error: 14.2210 - mean_absolute_percentage_error: 14.7053\n", "Epoch 144/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 536.9799 - mean_absolute_error: 14.0961 - mean_absolute_percentage_error: 14.5215\n", "Epoch 145/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 539.6394 - mean_absolute_error: 14.2113 - mean_absolute_percentage_error: 14.7146\n", "Epoch 146/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 538.7562 - mean_absolute_error: 14.0789 - mean_absolute_percentage_error: 14.5068\n", "Epoch 147/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 536.4084 - mean_absolute_error: 14.0958 - mean_absolute_percentage_error: 14.5496\n", "Epoch 148/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 539.0258 - mean_absolute_error: 14.2065 - mean_absolute_percentage_error: 14.7780\n", "Epoch 149/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 540.9120 - mean_absolute_error: 14.1493 - mean_absolute_percentage_error: 14.5889\n", "Epoch 150/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 533.8123 - mean_absolute_error: 14.1373 - mean_absolute_percentage_error: 14.6738\n", "Epoch 151/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 538.9818 - mean_absolute_error: 14.2818 - mean_absolute_percentage_error: 14.7994\n", "Epoch 152/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 535.3945 - mean_absolute_error: 14.1660 - mean_absolute_percentage_error: 14.6318\n", "Epoch 153/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 532.2208 - mean_absolute_error: 14.1532 - mean_absolute_percentage_error: 14.7001\n", "Epoch 154/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 527.9189 - mean_absolute_error: 14.1038 - mean_absolute_percentage_error: 14.6823\n", "Epoch 155/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 529.9041 - mean_absolute_error: 13.9457 - mean_absolute_percentage_error: 14.4363\n", "Epoch 156/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 527.2095 - mean_absolute_error: 13.9650 - mean_absolute_percentage_error: 14.4611\n", "Epoch 157/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 529.5710 - mean_absolute_error: 14.0640 - mean_absolute_percentage_error: 14.7056\n", "Epoch 158/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 521.7526 - mean_absolute_error: 14.0006 - mean_absolute_percentage_error: 14.6321\n", "Epoch 159/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 526.6138 - mean_absolute_error: 13.9618 - mean_absolute_percentage_error: 14.4909\n", "Epoch 160/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 524.5524 - mean_absolute_error: 14.0446 - mean_absolute_percentage_error: 14.6362\n", "Epoch 161/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 520.4467 - mean_absolute_error: 14.0361 - mean_absolute_percentage_error: 14.6886\n", "Epoch 162/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 521.0312 - mean_absolute_error: 13.9333 - mean_absolute_percentage_error: 14.5285\n", "Epoch 163/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 524.1859 - mean_absolute_error: 13.8301 - mean_absolute_percentage_error: 14.3494\n", "Epoch 164/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 520.4461 - mean_absolute_error: 13.8946 - mean_absolute_percentage_error: 14.4835\n", "Epoch 165/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 523.2239 - mean_absolute_error: 14.1263 - mean_absolute_percentage_error: 14.8474\n", "Epoch 166/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 521.7466 - mean_absolute_error: 13.9158 - mean_absolute_percentage_error: 14.5227\n", "Epoch 167/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 516.2388 - mean_absolute_error: 13.9417 - mean_absolute_percentage_error: 14.5941\n", "Epoch 168/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 524.2776 - mean_absolute_error: 13.9208 - mean_absolute_percentage_error: 14.4397\n", "Epoch 169/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 513.6591 - mean_absolute_error: 13.8267 - mean_absolute_percentage_error: 14.4166\n", "Epoch 170/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 514.7723 - mean_absolute_error: 13.9020 - mean_absolute_percentage_error: 14.5838\n", "Epoch 171/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 512.0550 - mean_absolute_error: 13.8438 - mean_absolute_percentage_error: 14.5337\n", "Epoch 172/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 509.9344 - mean_absolute_error: 13.8227 - mean_absolute_percentage_error: 14.4484\n", "Epoch 173/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 519.7784 - mean_absolute_error: 13.8312 - mean_absolute_percentage_error: 14.4077\n", "Epoch 174/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 511.0843 - mean_absolute_error: 13.8732 - mean_absolute_percentage_error: 14.5486\n", "Epoch 175/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 514.2950 - mean_absolute_error: 13.7550 - mean_absolute_percentage_error: 14.3617\n", "Epoch 176/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 514.0910 - mean_absolute_error: 13.9290 - mean_absolute_percentage_error: 14.6451\n", "Epoch 177/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 505.8664 - mean_absolute_error: 13.8389 - mean_absolute_percentage_error: 14.5599\n", "Epoch 178/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 509.5542 - mean_absolute_error: 13.6885 - mean_absolute_percentage_error: 14.2809\n", "Epoch 179/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 507.6328 - mean_absolute_error: 13.7776 - mean_absolute_percentage_error: 14.4559\n", "Epoch 180/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 509.7998 - mean_absolute_error: 13.8617 - mean_absolute_percentage_error: 14.5049\n", "Epoch 181/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 507.8986 - mean_absolute_error: 13.8280 - mean_absolute_percentage_error: 14.5308\n", "Epoch 182/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 510.6461 - mean_absolute_error: 13.7668 - mean_absolute_percentage_error: 14.4746\n", "Epoch 183/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 506.6362 - mean_absolute_error: 13.7056 - mean_absolute_percentage_error: 14.3469\n", "Epoch 184/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 504.4828 - mean_absolute_error: 13.7802 - mean_absolute_percentage_error: 14.4898\n", "Epoch 185/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 510.9891 - mean_absolute_error: 13.8467 - mean_absolute_percentage_error: 14.5503\n", "Epoch 186/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 502.2220 - mean_absolute_error: 13.8042 - mean_absolute_percentage_error: 14.5440\n", "Epoch 187/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 506.1675 - mean_absolute_error: 13.6575 - mean_absolute_percentage_error: 14.3165\n", "Epoch 188/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 499.2619 - mean_absolute_error: 13.6327 - mean_absolute_percentage_error: 14.3455\n", "Epoch 189/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 502.3392 - mean_absolute_error: 13.7972 - mean_absolute_percentage_error: 14.6003\n", "Epoch 190/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 500.2917 - mean_absolute_error: 13.6860 - mean_absolute_percentage_error: 14.3811\n", "Epoch 191/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 504.2755 - mean_absolute_error: 13.7246 - mean_absolute_percentage_error: 14.4657\n", "Epoch 192/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 502.6855 - mean_absolute_error: 13.6096 - mean_absolute_percentage_error: 14.2691\n", "Epoch 193/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 502.2716 - mean_absolute_error: 13.8658 - mean_absolute_percentage_error: 14.6619\n", "Epoch 194/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 497.2986 - mean_absolute_error: 13.6746 - mean_absolute_percentage_error: 14.4329\n", "Epoch 195/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 501.1582 - mean_absolute_error: 13.6982 - mean_absolute_percentage_error: 14.4180\n", "Epoch 196/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 492.5947 - mean_absolute_error: 13.5875 - mean_absolute_percentage_error: 14.3747\n", "Epoch 197/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 496.6996 - mean_absolute_error: 13.4928 - mean_absolute_percentage_error: 14.1782\n", "Epoch 198/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 502.5217 - mean_absolute_error: 13.8431 - mean_absolute_percentage_error: 14.6480\n", "Epoch 199/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 495.4518 - mean_absolute_error: 13.5663 - mean_absolute_percentage_error: 14.2806\n", "Epoch 200/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 490.9728 - mean_absolute_error: 13.6081 - mean_absolute_percentage_error: 14.3837\n", "Epoch 201/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 491.9117 - mean_absolute_error: 13.5025 - mean_absolute_percentage_error: 14.2449\n", "Epoch 202/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 491.8601 - mean_absolute_error: 13.5213 - mean_absolute_percentage_error: 14.2863\n", "Epoch 203/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 491.4577 - mean_absolute_error: 13.6068 - mean_absolute_percentage_error: 14.3880\n", "Epoch 204/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 489.6616 - mean_absolute_error: 13.5468 - mean_absolute_percentage_error: 14.3028\n", "Epoch 205/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 488.6859 - mean_absolute_error: 13.5211 - mean_absolute_percentage_error: 14.3096\n", "Epoch 206/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 488.5157 - mean_absolute_error: 13.5468 - mean_absolute_percentage_error: 14.3720\n", "Epoch 207/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 488.9674 - mean_absolute_error: 13.4820 - mean_absolute_percentage_error: 14.2032\n", "Epoch 208/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 487.4901 - mean_absolute_error: 13.4367 - mean_absolute_percentage_error: 14.1931\n", "Epoch 209/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 487.8140 - mean_absolute_error: 13.5175 - mean_absolute_percentage_error: 14.3164\n", "Epoch 210/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 486.8511 - mean_absolute_error: 13.5534 - mean_absolute_percentage_error: 14.3556\n", "Epoch 211/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 487.5799 - mean_absolute_error: 13.4882 - mean_absolute_percentage_error: 14.2955\n", "Epoch 212/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 482.9704 - mean_absolute_error: 13.4363 - mean_absolute_percentage_error: 14.2665\n", "Epoch 213/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 487.9311 - mean_absolute_error: 13.5707 - mean_absolute_percentage_error: 14.3647\n", "Epoch 214/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 483.3335 - mean_absolute_error: 13.4671 - mean_absolute_percentage_error: 14.2766\n", "Epoch 215/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 482.4341 - mean_absolute_error: 13.4484 - mean_absolute_percentage_error: 14.2813\n", "Epoch 216/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 486.4435 - mean_absolute_error: 13.4856 - mean_absolute_percentage_error: 14.3094\n", "Epoch 217/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 483.8159 - mean_absolute_error: 13.4384 - mean_absolute_percentage_error: 14.1906\n", "Epoch 218/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 484.4789 - mean_absolute_error: 13.5253 - mean_absolute_percentage_error: 14.3352\n", "Epoch 219/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 488.4053 - mean_absolute_error: 13.6378 - mean_absolute_percentage_error: 14.5217\n", "Epoch 220/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 485.7378 - mean_absolute_error: 13.3894 - mean_absolute_percentage_error: 14.1673\n", "Epoch 221/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 480.6607 - mean_absolute_error: 13.4315 - mean_absolute_percentage_error: 14.2042\n", "Epoch 222/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 481.3052 - mean_absolute_error: 13.4222 - mean_absolute_percentage_error: 14.2750\n", "Epoch 223/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 480.1549 - mean_absolute_error: 13.4936 - mean_absolute_percentage_error: 14.3892\n", "Epoch 224/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 480.0995 - mean_absolute_error: 13.3205 - mean_absolute_percentage_error: 14.1229\n", "Epoch 225/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 479.6157 - mean_absolute_error: 13.3809 - mean_absolute_percentage_error: 14.2531\n", "Epoch 226/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 480.6735 - mean_absolute_error: 13.4617 - mean_absolute_percentage_error: 14.3441\n", "Epoch 227/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 476.1359 - mean_absolute_error: 13.3921 - mean_absolute_percentage_error: 14.2747\n", "Epoch 228/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 482.4225 - mean_absolute_error: 13.4886 - mean_absolute_percentage_error: 14.2707\n", "Epoch 229/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 474.5368 - mean_absolute_error: 13.3294 - mean_absolute_percentage_error: 14.1922\n", "Epoch 230/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 478.1674 - mean_absolute_error: 13.3848 - mean_absolute_percentage_error: 14.1720\n", "Epoch 231/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 477.4900 - mean_absolute_error: 13.3564 - mean_absolute_percentage_error: 14.2201\n", "Epoch 232/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 475.7230 - mean_absolute_error: 13.3765 - mean_absolute_percentage_error: 14.2367\n", "Epoch 233/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 474.1965 - mean_absolute_error: 13.4051 - mean_absolute_percentage_error: 14.2777\n", "Epoch 234/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 475.6379 - mean_absolute_error: 13.3880 - mean_absolute_percentage_error: 14.3288\n", "Epoch 235/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 473.3627 - mean_absolute_error: 13.2724 - mean_absolute_percentage_error: 14.1307\n", "Epoch 236/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 474.4074 - mean_absolute_error: 13.3044 - mean_absolute_percentage_error: 14.2038\n", "Epoch 237/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 474.2868 - mean_absolute_error: 13.2960 - mean_absolute_percentage_error: 14.1492\n", "Epoch 238/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 471.1841 - mean_absolute_error: 13.2512 - mean_absolute_percentage_error: 14.1449\n", "Epoch 239/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 471.1591 - mean_absolute_error: 13.2817 - mean_absolute_percentage_error: 14.1670\n", "Epoch 240/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 472.5487 - mean_absolute_error: 13.2102 - mean_absolute_percentage_error: 14.0486\n", "Epoch 241/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 469.1688 - mean_absolute_error: 13.2911 - mean_absolute_percentage_error: 14.2663\n", "Epoch 242/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 471.4414 - mean_absolute_error: 13.2927 - mean_absolute_percentage_error: 14.2662\n", "Epoch 243/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 475.4010 - mean_absolute_error: 13.2879 - mean_absolute_percentage_error: 14.0722\n", "Epoch 244/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 474.5285 - mean_absolute_error: 13.3866 - mean_absolute_percentage_error: 14.3001\n", "Epoch 245/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 469.8882 - mean_absolute_error: 13.2202 - mean_absolute_percentage_error: 14.1867\n", "Epoch 246/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 469.9311 - mean_absolute_error: 13.2102 - mean_absolute_percentage_error: 14.0324\n", "Epoch 247/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 469.0129 - mean_absolute_error: 13.3110 - mean_absolute_percentage_error: 14.2606\n", "Epoch 248/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 465.0145 - mean_absolute_error: 13.1724 - mean_absolute_percentage_error: 14.1394\n", "Epoch 249/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 465.9364 - mean_absolute_error: 13.1652 - mean_absolute_percentage_error: 14.0765\n", "Epoch 250/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 465.3367 - mean_absolute_error: 13.1714 - mean_absolute_percentage_error: 14.1022\n", "Epoch 251/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 466.8428 - mean_absolute_error: 13.1839 - mean_absolute_percentage_error: 14.0994\n", "Epoch 252/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 463.4367 - mean_absolute_error: 13.1041 - mean_absolute_percentage_error: 14.0417\n", "Epoch 253/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 464.2352 - mean_absolute_error: 13.2262 - mean_absolute_percentage_error: 14.1903\n", "Epoch 254/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 465.0757 - mean_absolute_error: 13.1377 - mean_absolute_percentage_error: 14.0708\n", "Epoch 255/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 463.5256 - mean_absolute_error: 13.0957 - mean_absolute_percentage_error: 14.0307\n", "Epoch 256/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 463.8532 - mean_absolute_error: 13.1673 - mean_absolute_percentage_error: 14.1558\n", "Epoch 257/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 462.5304 - mean_absolute_error: 13.1277 - mean_absolute_percentage_error: 14.0767\n", "Epoch 258/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 464.2828 - mean_absolute_error: 13.1053 - mean_absolute_percentage_error: 14.0456\n", "Epoch 259/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 463.8347 - mean_absolute_error: 13.1236 - mean_absolute_percentage_error: 14.0864\n", "Epoch 260/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 460.8294 - mean_absolute_error: 13.0817 - mean_absolute_percentage_error: 14.0730\n", "Epoch 261/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 460.3183 - mean_absolute_error: 13.1213 - mean_absolute_percentage_error: 14.0832\n", "Epoch 262/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 460.1303 - mean_absolute_error: 13.1219 - mean_absolute_percentage_error: 14.0826\n", "Epoch 263/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 461.2187 - mean_absolute_error: 13.1692 - mean_absolute_percentage_error: 14.1505\n", "Epoch 264/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 461.0131 - mean_absolute_error: 13.0866 - mean_absolute_percentage_error: 14.0100\n", "Epoch 265/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 461.2482 - mean_absolute_error: 13.0527 - mean_absolute_percentage_error: 13.9929\n", "Epoch 266/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 458.9508 - mean_absolute_error: 13.0898 - mean_absolute_percentage_error: 14.0857\n", "Epoch 267/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 457.5953 - mean_absolute_error: 13.0183 - mean_absolute_percentage_error: 14.0154\n", "Epoch 268/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 458.7173 - mean_absolute_error: 13.0352 - mean_absolute_percentage_error: 14.0105\n", "Epoch 269/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 459.0412 - mean_absolute_error: 13.1094 - mean_absolute_percentage_error: 14.0671\n", "Epoch 270/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 458.5614 - mean_absolute_error: 13.0919 - mean_absolute_percentage_error: 14.1119\n", "Epoch 271/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 459.1760 - mean_absolute_error: 13.1545 - mean_absolute_percentage_error: 14.1357\n", "Epoch 272/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 455.3437 - mean_absolute_error: 13.0390 - mean_absolute_percentage_error: 14.0343\n", "Epoch 273/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 459.9648 - mean_absolute_error: 13.0930 - mean_absolute_percentage_error: 14.1059\n", "Epoch 274/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 458.1582 - mean_absolute_error: 13.0238 - mean_absolute_percentage_error: 13.9983\n", "Epoch 275/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 457.1316 - mean_absolute_error: 13.0506 - mean_absolute_percentage_error: 14.0686\n", "Epoch 276/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 458.1903 - mean_absolute_error: 13.0364 - mean_absolute_percentage_error: 14.0019\n", "Epoch 277/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 457.3430 - mean_absolute_error: 13.0549 - mean_absolute_percentage_error: 14.0786\n", "Epoch 278/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 459.3451 - mean_absolute_error: 13.1003 - mean_absolute_percentage_error: 14.1421\n", "Epoch 279/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 455.4984 - mean_absolute_error: 13.0730 - mean_absolute_percentage_error: 14.0881\n", "Epoch 280/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 453.0569 - mean_absolute_error: 12.9585 - mean_absolute_percentage_error: 13.9711\n", "Epoch 281/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 457.5428 - mean_absolute_error: 13.0291 - mean_absolute_percentage_error: 14.0431\n", "Epoch 282/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 455.2252 - mean_absolute_error: 12.9605 - mean_absolute_percentage_error: 13.9310\n", "Epoch 283/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 454.3318 - mean_absolute_error: 12.9594 - mean_absolute_percentage_error: 13.9719\n", "Epoch 284/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 453.9716 - mean_absolute_error: 13.0025 - mean_absolute_percentage_error: 14.0200\n", "Epoch 285/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 455.2581 - mean_absolute_error: 13.1353 - mean_absolute_percentage_error: 14.1193\n", "Epoch 286/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 456.3631 - mean_absolute_error: 13.1130 - mean_absolute_percentage_error: 14.1582\n", "Epoch 287/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 456.5887 - mean_absolute_error: 12.9812 - mean_absolute_percentage_error: 13.9633\n", "Epoch 288/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 453.3662 - mean_absolute_error: 12.9368 - mean_absolute_percentage_error: 13.9314\n", "Epoch 289/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 451.7266 - mean_absolute_error: 13.0376 - mean_absolute_percentage_error: 14.0929\n", "Epoch 290/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 451.4980 - mean_absolute_error: 12.9270 - mean_absolute_percentage_error: 13.9771\n", "Epoch 291/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 453.4079 - mean_absolute_error: 12.9649 - mean_absolute_percentage_error: 13.9584\n", "Epoch 292/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 450.7113 - mean_absolute_error: 13.0173 - mean_absolute_percentage_error: 14.0568\n", "Epoch 293/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 453.3527 - mean_absolute_error: 13.0041 - mean_absolute_percentage_error: 14.0359\n", "Epoch 294/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 455.0202 - mean_absolute_error: 12.9426 - mean_absolute_percentage_error: 13.8717\n", "Epoch 295/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 452.1084 - mean_absolute_error: 13.0294 - mean_absolute_percentage_error: 14.1141\n", "Epoch 296/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 450.2031 - mean_absolute_error: 12.9898 - mean_absolute_percentage_error: 14.0703\n", "Epoch 297/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 450.1469 - mean_absolute_error: 12.8881 - mean_absolute_percentage_error: 13.8528\n", "Epoch 298/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 449.2065 - mean_absolute_error: 12.9134 - mean_absolute_percentage_error: 13.9736\n", "Epoch 299/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 447.8001 - mean_absolute_error: 12.9379 - mean_absolute_percentage_error: 14.0371\n", "Epoch 300/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 447.9913 - mean_absolute_error: 12.8957 - mean_absolute_percentage_error: 13.9507\n", "Epoch 301/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 447.8429 - mean_absolute_error: 12.8449 - mean_absolute_percentage_error: 13.8730\n", "Epoch 302/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 447.4172 - mean_absolute_error: 12.8703 - mean_absolute_percentage_error: 13.9291\n", "Epoch 303/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 451.2496 - mean_absolute_error: 12.9857 - mean_absolute_percentage_error: 14.0188\n", "Epoch 304/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 447.7301 - mean_absolute_error: 13.0067 - mean_absolute_percentage_error: 14.1122\n", "Epoch 305/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 449.2870 - mean_absolute_error: 12.9413 - mean_absolute_percentage_error: 13.9958\n", "Epoch 306/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 446.4304 - mean_absolute_error: 12.8781 - mean_absolute_percentage_error: 13.9291\n", "Epoch 307/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 447.3802 - mean_absolute_error: 12.9282 - mean_absolute_percentage_error: 14.0166\n", "Epoch 308/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 447.2425 - mean_absolute_error: 12.9696 - mean_absolute_percentage_error: 14.0900\n", "Epoch 309/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 445.3722 - mean_absolute_error: 12.8477 - mean_absolute_percentage_error: 13.8991\n", "Epoch 310/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 445.0899 - mean_absolute_error: 12.8115 - mean_absolute_percentage_error: 13.8858\n", "Epoch 311/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 446.3776 - mean_absolute_error: 12.8766 - mean_absolute_percentage_error: 13.9777\n", "Epoch 312/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 443.5902 - mean_absolute_error: 12.7866 - mean_absolute_percentage_error: 13.8495\n", "Epoch 313/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 444.5401 - mean_absolute_error: 12.7851 - mean_absolute_percentage_error: 13.8450\n", "Epoch 314/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 445.7879 - mean_absolute_error: 12.8841 - mean_absolute_percentage_error: 13.9858\n", "Epoch 315/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 444.8242 - mean_absolute_error: 12.8848 - mean_absolute_percentage_error: 13.9887\n", "Epoch 316/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 444.2825 - mean_absolute_error: 12.8173 - mean_absolute_percentage_error: 13.9074\n", "Epoch 317/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 444.0893 - mean_absolute_error: 12.8545 - mean_absolute_percentage_error: 13.9804\n", "Epoch 318/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 444.8220 - mean_absolute_error: 12.8647 - mean_absolute_percentage_error: 13.9342\n", "Epoch 319/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 442.7452 - mean_absolute_error: 12.8616 - mean_absolute_percentage_error: 13.9592\n", "Epoch 320/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 444.2711 - mean_absolute_error: 12.7844 - mean_absolute_percentage_error: 13.8677\n", "Epoch 321/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 443.5755 - mean_absolute_error: 12.7971 - mean_absolute_percentage_error: 13.8770\n", "Epoch 322/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 442.1266 - mean_absolute_error: 12.8396 - mean_absolute_percentage_error: 13.9209\n", "Epoch 323/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 447.2940 - mean_absolute_error: 12.9235 - mean_absolute_percentage_error: 13.9896\n", "Epoch 324/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 444.7495 - mean_absolute_error: 12.8477 - mean_absolute_percentage_error: 13.9248\n", "Epoch 325/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 443.2931 - mean_absolute_error: 12.8409 - mean_absolute_percentage_error: 13.9095\n", "Epoch 326/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 441.9189 - mean_absolute_error: 12.8270 - mean_absolute_percentage_error: 13.9098\n", "Epoch 327/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 439.4569 - mean_absolute_error: 12.7003 - mean_absolute_percentage_error: 13.8161\n", "Epoch 328/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 439.0885 - mean_absolute_error: 12.7466 - mean_absolute_percentage_error: 13.8598\n", "Epoch 329/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 438.5906 - mean_absolute_error: 12.7342 - mean_absolute_percentage_error: 13.8404\n", "Epoch 330/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 439.2990 - mean_absolute_error: 12.7630 - mean_absolute_percentage_error: 13.8555\n", "Epoch 331/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 438.1302 - mean_absolute_error: 12.7352 - mean_absolute_percentage_error: 13.8254\n", "Epoch 332/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 438.3800 - mean_absolute_error: 12.6578 - mean_absolute_percentage_error: 13.7546\n", "Epoch 333/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 440.8786 - mean_absolute_error: 12.7310 - mean_absolute_percentage_error: 13.8505\n", "Epoch 334/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 438.1957 - mean_absolute_error: 12.6977 - mean_absolute_percentage_error: 13.8108\n", "Epoch 335/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 438.2702 - mean_absolute_error: 12.7412 - mean_absolute_percentage_error: 13.8417\n", "Epoch 336/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 442.0157 - mean_absolute_error: 12.8271 - mean_absolute_percentage_error: 13.9229\n", "Epoch 337/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 438.6081 - mean_absolute_error: 12.7765 - mean_absolute_percentage_error: 13.8845\n", "Epoch 338/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 438.0249 - mean_absolute_error: 12.7507 - mean_absolute_percentage_error: 13.8611\n", "Epoch 339/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 437.3978 - mean_absolute_error: 12.6553 - mean_absolute_percentage_error: 13.7417\n", "Epoch 340/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 435.8968 - mean_absolute_error: 12.7067 - mean_absolute_percentage_error: 13.8517\n", "Epoch 341/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 436.7429 - mean_absolute_error: 12.6671 - mean_absolute_percentage_error: 13.7681\n", "Epoch 342/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 436.4738 - mean_absolute_error: 12.6823 - mean_absolute_percentage_error: 13.8122\n", "Epoch 343/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 435.2958 - mean_absolute_error: 12.6366 - mean_absolute_percentage_error: 13.7248\n", "Epoch 344/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 436.9108 - mean_absolute_error: 12.7178 - mean_absolute_percentage_error: 13.8277\n", "Epoch 345/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 436.7537 - mean_absolute_error: 12.7484 - mean_absolute_percentage_error: 13.8908\n", "Epoch 346/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 437.0537 - mean_absolute_error: 12.7343 - mean_absolute_percentage_error: 13.8733\n", "Epoch 347/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 433.9487 - mean_absolute_error: 12.6641 - mean_absolute_percentage_error: 13.8424\n", "Epoch 348/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 436.3140 - mean_absolute_error: 12.6105 - mean_absolute_percentage_error: 13.6622\n", "Epoch 349/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 436.1429 - mean_absolute_error: 12.7079 - mean_absolute_percentage_error: 13.8722\n", "Epoch 350/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 434.3533 - mean_absolute_error: 12.6799 - mean_absolute_percentage_error: 13.8381\n", "Epoch 351/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 435.5850 - mean_absolute_error: 12.6653 - mean_absolute_percentage_error: 13.7649\n", "Epoch 352/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 434.2619 - mean_absolute_error: 12.6447 - mean_absolute_percentage_error: 13.7241\n", "Epoch 353/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 435.5435 - mean_absolute_error: 12.6680 - mean_absolute_percentage_error: 13.8378\n", "Epoch 354/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 433.3014 - mean_absolute_error: 12.6780 - mean_absolute_percentage_error: 13.8189\n", "Epoch 355/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 433.8479 - mean_absolute_error: 12.6301 - mean_absolute_percentage_error: 13.7702\n", "Epoch 356/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 433.6459 - mean_absolute_error: 12.6089 - mean_absolute_percentage_error: 13.7583\n", "Epoch 357/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 432.0153 - mean_absolute_error: 12.6308 - mean_absolute_percentage_error: 13.7703\n", "Epoch 358/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 433.6544 - mean_absolute_error: 12.5922 - mean_absolute_percentage_error: 13.6976\n", "Epoch 359/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 430.9907 - mean_absolute_error: 12.5504 - mean_absolute_percentage_error: 13.6526\n", "Epoch 360/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 431.8987 - mean_absolute_error: 12.6419 - mean_absolute_percentage_error: 13.8471\n", "Epoch 361/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 431.7182 - mean_absolute_error: 12.6317 - mean_absolute_percentage_error: 13.7813\n", "Epoch 362/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 430.1259 - mean_absolute_error: 12.5566 - mean_absolute_percentage_error: 13.6961\n", "Epoch 363/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 432.2221 - mean_absolute_error: 12.6068 - mean_absolute_percentage_error: 13.7358\n", "Epoch 364/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 428.9785 - mean_absolute_error: 12.5082 - mean_absolute_percentage_error: 13.6833\n", "Epoch 365/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 431.5387 - mean_absolute_error: 12.5770 - mean_absolute_percentage_error: 13.7239\n", "Epoch 366/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 430.5910 - mean_absolute_error: 12.5734 - mean_absolute_percentage_error: 13.7566\n", "Epoch 367/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 430.6709 - mean_absolute_error: 12.5424 - mean_absolute_percentage_error: 13.6832\n", "Epoch 368/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 429.9673 - mean_absolute_error: 12.6281 - mean_absolute_percentage_error: 13.7996\n", "Epoch 369/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 431.1599 - mean_absolute_error: 12.6542 - mean_absolute_percentage_error: 13.8036\n", "Epoch 370/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 431.5043 - mean_absolute_error: 12.5688 - mean_absolute_percentage_error: 13.6727\n", "Epoch 371/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 429.8430 - mean_absolute_error: 12.6260 - mean_absolute_percentage_error: 13.8119\n", "Epoch 372/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 428.7260 - mean_absolute_error: 12.5253 - mean_absolute_percentage_error: 13.7086\n", "Epoch 373/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 429.3631 - mean_absolute_error: 12.5412 - mean_absolute_percentage_error: 13.6681\n", "Epoch 374/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 430.0153 - mean_absolute_error: 12.6394 - mean_absolute_percentage_error: 13.8046\n", "Epoch 375/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 430.6500 - mean_absolute_error: 12.5773 - mean_absolute_percentage_error: 13.7647\n", "Epoch 376/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 429.6545 - mean_absolute_error: 12.5399 - mean_absolute_percentage_error: 13.6467\n", "Epoch 377/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 428.5045 - mean_absolute_error: 12.5244 - mean_absolute_percentage_error: 13.6650\n", "Epoch 378/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 430.3684 - mean_absolute_error: 12.6076 - mean_absolute_percentage_error: 13.7748\n", "Epoch 379/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 428.7687 - mean_absolute_error: 12.5780 - mean_absolute_percentage_error: 13.7513\n", "Epoch 380/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 426.8940 - mean_absolute_error: 12.5416 - mean_absolute_percentage_error: 13.7005\n", "Epoch 381/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 428.8385 - mean_absolute_error: 12.5486 - mean_absolute_percentage_error: 13.7109\n", "Epoch 382/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 428.4724 - mean_absolute_error: 12.4902 - mean_absolute_percentage_error: 13.6431\n", "Epoch 383/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 428.0958 - mean_absolute_error: 12.5712 - mean_absolute_percentage_error: 13.7510\n", "Epoch 384/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 427.5721 - mean_absolute_error: 12.5789 - mean_absolute_percentage_error: 13.7807\n", "Epoch 385/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 426.5889 - mean_absolute_error: 12.5333 - mean_absolute_percentage_error: 13.7104\n", "Epoch 386/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 426.8146 - mean_absolute_error: 12.5390 - mean_absolute_percentage_error: 13.7198\n", "Epoch 387/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 425.3283 - mean_absolute_error: 12.4595 - mean_absolute_percentage_error: 13.6343\n", "Epoch 388/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 426.2201 - mean_absolute_error: 12.4780 - mean_absolute_percentage_error: 13.6754\n", "Epoch 389/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 426.8235 - mean_absolute_error: 12.5488 - mean_absolute_percentage_error: 13.7275\n", "Epoch 390/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 425.8202 - mean_absolute_error: 12.5239 - mean_absolute_percentage_error: 13.7243\n", "Epoch 391/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 427.8373 - mean_absolute_error: 12.5091 - mean_absolute_percentage_error: 13.6400\n", "Epoch 392/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 426.8843 - mean_absolute_error: 12.5400 - mean_absolute_percentage_error: 13.7147\n", "Epoch 393/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 425.1960 - mean_absolute_error: 12.5055 - mean_absolute_percentage_error: 13.7342\n", "Epoch 394/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 424.9663 - mean_absolute_error: 12.4574 - mean_absolute_percentage_error: 13.6229\n", "Epoch 395/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 428.0396 - mean_absolute_error: 12.5777 - mean_absolute_percentage_error: 13.7344\n", "Epoch 396/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 426.6123 - mean_absolute_error: 12.5082 - mean_absolute_percentage_error: 13.7171\n", "Epoch 397/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 427.4333 - mean_absolute_error: 12.5231 - mean_absolute_percentage_error: 13.6171\n", "Epoch 398/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 427.2609 - mean_absolute_error: 12.6111 - mean_absolute_percentage_error: 13.8279\n", "Epoch 399/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 423.8147 - mean_absolute_error: 12.4964 - mean_absolute_percentage_error: 13.6699\n", "Epoch 400/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 423.2380 - mean_absolute_error: 12.4961 - mean_absolute_percentage_error: 13.7077\n", "Epoch 401/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 424.5359 - mean_absolute_error: 12.4634 - mean_absolute_percentage_error: 13.6441\n", "Epoch 402/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 422.7496 - mean_absolute_error: 12.4545 - mean_absolute_percentage_error: 13.6679\n", "Epoch 403/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 422.2954 - mean_absolute_error: 12.4322 - mean_absolute_percentage_error: 13.6660\n", "Epoch 404/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 423.7608 - mean_absolute_error: 12.4923 - mean_absolute_percentage_error: 13.6679\n", "Epoch 405/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 421.7570 - mean_absolute_error: 12.4354 - mean_absolute_percentage_error: 13.6202\n", "Epoch 406/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 423.2868 - mean_absolute_error: 12.4318 - mean_absolute_percentage_error: 13.6297\n", "Epoch 407/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 424.9512 - mean_absolute_error: 12.5142 - mean_absolute_percentage_error: 13.7182\n", "Epoch 408/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 422.4470 - mean_absolute_error: 12.5033 - mean_absolute_percentage_error: 13.7101\n", "Epoch 409/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 421.7734 - mean_absolute_error: 12.4340 - mean_absolute_percentage_error: 13.6178\n", "Epoch 410/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 422.2756 - mean_absolute_error: 12.4232 - mean_absolute_percentage_error: 13.6512\n", "Epoch 411/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 421.9445 - mean_absolute_error: 12.4832 - mean_absolute_percentage_error: 13.6535\n", "Epoch 412/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 423.9546 - mean_absolute_error: 12.4652 - mean_absolute_percentage_error: 13.6246\n", "Epoch 413/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 425.1797 - mean_absolute_error: 12.5600 - mean_absolute_percentage_error: 13.6764\n", "Epoch 414/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 424.0365 - mean_absolute_error: 12.5076 - mean_absolute_percentage_error: 13.7024\n", "Epoch 415/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 420.2200 - mean_absolute_error: 12.4262 - mean_absolute_percentage_error: 13.6733\n", "Epoch 416/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 420.2119 - mean_absolute_error: 12.3666 - mean_absolute_percentage_error: 13.5711\n", "Epoch 417/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 423.2134 - mean_absolute_error: 12.4701 - mean_absolute_percentage_error: 13.6584\n", "Epoch 418/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 420.6377 - mean_absolute_error: 12.4305 - mean_absolute_percentage_error: 13.6281\n", "Epoch 419/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 420.0645 - mean_absolute_error: 12.3911 - mean_absolute_percentage_error: 13.5946\n", "Epoch 420/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 420.7085 - mean_absolute_error: 12.4435 - mean_absolute_percentage_error: 13.6736\n", "Epoch 421/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 420.2659 - mean_absolute_error: 12.4292 - mean_absolute_percentage_error: 13.6203\n", "Epoch 422/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 420.4348 - mean_absolute_error: 12.4347 - mean_absolute_percentage_error: 13.6460\n", "Epoch 423/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 420.0414 - mean_absolute_error: 12.4073 - mean_absolute_percentage_error: 13.6256\n", "Epoch 424/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 418.5494 - mean_absolute_error: 12.3497 - mean_absolute_percentage_error: 13.5608\n", "Epoch 425/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 421.6267 - mean_absolute_error: 12.4766 - mean_absolute_percentage_error: 13.6967\n", "Epoch 426/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 418.6843 - mean_absolute_error: 12.3957 - mean_absolute_percentage_error: 13.5864\n", "Epoch 427/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 420.4724 - mean_absolute_error: 12.3756 - mean_absolute_percentage_error: 13.5443\n", "Epoch 428/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 419.1703 - mean_absolute_error: 12.4139 - mean_absolute_percentage_error: 13.6605\n", "Epoch 429/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 418.8851 - mean_absolute_error: 12.4482 - mean_absolute_percentage_error: 13.6462\n", "Epoch 430/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 417.7826 - mean_absolute_error: 12.4217 - mean_absolute_percentage_error: 13.6625\n", "Epoch 431/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 417.6522 - mean_absolute_error: 12.3929 - mean_absolute_percentage_error: 13.6190\n", "Epoch 432/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 421.1663 - mean_absolute_error: 12.4228 - mean_absolute_percentage_error: 13.5755\n", "Epoch 433/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 419.2278 - mean_absolute_error: 12.3875 - mean_absolute_percentage_error: 13.6303\n", "Epoch 434/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 418.0834 - mean_absolute_error: 12.3891 - mean_absolute_percentage_error: 13.5986\n", "Epoch 435/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 417.5318 - mean_absolute_error: 12.3294 - mean_absolute_percentage_error: 13.5494\n", "Epoch 436/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 418.4147 - mean_absolute_error: 12.4081 - mean_absolute_percentage_error: 13.6127\n", "Epoch 437/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 418.2980 - mean_absolute_error: 12.4034 - mean_absolute_percentage_error: 13.6075\n", "Epoch 438/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 418.2492 - mean_absolute_error: 12.3112 - mean_absolute_percentage_error: 13.4957\n", "Epoch 439/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 417.2125 - mean_absolute_error: 12.3879 - mean_absolute_percentage_error: 13.6455\n", "Epoch 440/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 418.2238 - mean_absolute_error: 12.4602 - mean_absolute_percentage_error: 13.7215\n", "Epoch 441/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 416.7845 - mean_absolute_error: 12.3849 - mean_absolute_percentage_error: 13.6211\n", "Epoch 442/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 417.1416 - mean_absolute_error: 12.3558 - mean_absolute_percentage_error: 13.5457\n", "Epoch 443/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 417.4982 - mean_absolute_error: 12.3396 - mean_absolute_percentage_error: 13.4951\n", "Epoch 444/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 418.1409 - mean_absolute_error: 12.4282 - mean_absolute_percentage_error: 13.6987\n", "Epoch 445/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 417.1277 - mean_absolute_error: 12.3745 - mean_absolute_percentage_error: 13.6105\n", "Epoch 446/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 415.4851 - mean_absolute_error: 12.3223 - mean_absolute_percentage_error: 13.5168\n", "Epoch 447/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 415.7296 - mean_absolute_error: 12.3336 - mean_absolute_percentage_error: 13.5574\n", "Epoch 448/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 418.3037 - mean_absolute_error: 12.4346 - mean_absolute_percentage_error: 13.6226\n", "Epoch 449/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 418.5910 - mean_absolute_error: 12.4280 - mean_absolute_percentage_error: 13.6818\n", "Epoch 450/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 415.5983 - mean_absolute_error: 12.3897 - mean_absolute_percentage_error: 13.6328\n", "Epoch 451/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 417.6087 - mean_absolute_error: 12.4048 - mean_absolute_percentage_error: 13.5888\n", "Epoch 452/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 415.2840 - mean_absolute_error: 12.3341 - mean_absolute_percentage_error: 13.5410\n", "Epoch 453/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 417.3434 - mean_absolute_error: 12.4279 - mean_absolute_percentage_error: 13.6781\n", "Epoch 454/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 415.6289 - mean_absolute_error: 12.3715 - mean_absolute_percentage_error: 13.6194\n", "Epoch 455/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 415.9023 - mean_absolute_error: 12.3205 - mean_absolute_percentage_error: 13.4731\n", "Epoch 456/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 414.2253 - mean_absolute_error: 12.3086 - mean_absolute_percentage_error: 13.5232\n", "Epoch 457/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 415.3801 - mean_absolute_error: 12.3589 - mean_absolute_percentage_error: 13.5877\n", "Epoch 458/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 414.1558 - mean_absolute_error: 12.3128 - mean_absolute_percentage_error: 13.5277\n", "Epoch 459/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 413.9081 - mean_absolute_error: 12.3407 - mean_absolute_percentage_error: 13.5967\n", "Epoch 460/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 414.8241 - mean_absolute_error: 12.3936 - mean_absolute_percentage_error: 13.6225\n", "Epoch 461/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 414.8943 - mean_absolute_error: 12.3461 - mean_absolute_percentage_error: 13.6106\n", "Epoch 462/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 414.4001 - mean_absolute_error: 12.3156 - mean_absolute_percentage_error: 13.5190\n", "Epoch 463/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 413.3298 - mean_absolute_error: 12.2984 - mean_absolute_percentage_error: 13.5317\n", "Epoch 464/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 414.7062 - mean_absolute_error: 12.3378 - mean_absolute_percentage_error: 13.5921\n", "Epoch 465/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 413.9068 - mean_absolute_error: 12.3258 - mean_absolute_percentage_error: 13.5783\n", "Epoch 466/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 414.0020 - mean_absolute_error: 12.3729 - mean_absolute_percentage_error: 13.5676\n", "Epoch 467/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 413.8453 - mean_absolute_error: 12.3738 - mean_absolute_percentage_error: 13.6288\n", "Epoch 468/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 413.3268 - mean_absolute_error: 12.3260 - mean_absolute_percentage_error: 13.6078\n", "Epoch 469/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 413.1558 - mean_absolute_error: 12.3177 - mean_absolute_percentage_error: 13.5276\n", "Epoch 470/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 413.4508 - mean_absolute_error: 12.3179 - mean_absolute_percentage_error: 13.5797\n", "Epoch 471/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 412.9333 - mean_absolute_error: 12.2825 - mean_absolute_percentage_error: 13.5380\n", "Epoch 472/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 412.3635 - mean_absolute_error: 12.3289 - mean_absolute_percentage_error: 13.5632\n", "Epoch 473/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 411.1367 - mean_absolute_error: 12.2829 - mean_absolute_percentage_error: 13.5469\n", "Epoch 474/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 411.6473 - mean_absolute_error: 12.2893 - mean_absolute_percentage_error: 13.5477\n", "Epoch 475/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 411.2452 - mean_absolute_error: 12.2579 - mean_absolute_percentage_error: 13.5111\n", "Epoch 476/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 412.2820 - mean_absolute_error: 12.2921 - mean_absolute_percentage_error: 13.5528\n", "Epoch 477/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 411.7474 - mean_absolute_error: 12.2591 - mean_absolute_percentage_error: 13.4880\n", "Epoch 478/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 412.1250 - mean_absolute_error: 12.3388 - mean_absolute_percentage_error: 13.5919\n", "Epoch 479/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 410.9256 - mean_absolute_error: 12.2946 - mean_absolute_percentage_error: 13.5615\n", "Epoch 480/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 414.2627 - mean_absolute_error: 12.3439 - mean_absolute_percentage_error: 13.5606\n", "Epoch 481/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 412.4510 - mean_absolute_error: 12.3476 - mean_absolute_percentage_error: 13.5745\n", "Epoch 482/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 411.7901 - mean_absolute_error: 12.3164 - mean_absolute_percentage_error: 13.5817\n", "Epoch 483/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 412.1971 - mean_absolute_error: 12.3384 - mean_absolute_percentage_error: 13.5785\n", "Epoch 484/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 412.0933 - mean_absolute_error: 12.3343 - mean_absolute_percentage_error: 13.5472\n", "Epoch 485/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 411.4384 - mean_absolute_error: 12.2753 - mean_absolute_percentage_error: 13.5264\n", "Epoch 486/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 410.1884 - mean_absolute_error: 12.2954 - mean_absolute_percentage_error: 13.5655\n", "Epoch 487/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 411.3309 - mean_absolute_error: 12.2700 - mean_absolute_percentage_error: 13.4929\n", "Epoch 488/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 412.6119 - mean_absolute_error: 12.3328 - mean_absolute_percentage_error: 13.6021\n", "Epoch 489/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 411.8277 - mean_absolute_error: 12.3612 - mean_absolute_percentage_error: 13.6321\n", "Epoch 490/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 410.5264 - mean_absolute_error: 12.2913 - mean_absolute_percentage_error: 13.5486\n", "Epoch 491/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 410.6336 - mean_absolute_error: 12.2803 - mean_absolute_percentage_error: 13.5190\n", "Epoch 492/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.9079 - mean_absolute_error: 12.2619 - mean_absolute_percentage_error: 13.5017\n", "Epoch 493/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.4432 - mean_absolute_error: 12.2476 - mean_absolute_percentage_error: 13.5377\n", "Epoch 494/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 408.9451 - mean_absolute_error: 12.2617 - mean_absolute_percentage_error: 13.5495\n", "Epoch 495/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.1828 - mean_absolute_error: 12.2404 - mean_absolute_percentage_error: 13.5197\n", "Epoch 496/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.3631 - mean_absolute_error: 12.2317 - mean_absolute_percentage_error: 13.4692\n", "Epoch 497/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 408.5021 - mean_absolute_error: 12.2471 - mean_absolute_percentage_error: 13.5245\n", "Epoch 498/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.4647 - mean_absolute_error: 12.2636 - mean_absolute_percentage_error: 13.5549\n", "Epoch 499/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 408.1298 - mean_absolute_error: 12.2211 - mean_absolute_percentage_error: 13.4620\n", "Epoch 500/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.0601 - mean_absolute_error: 12.2184 - mean_absolute_percentage_error: 13.4928\n", "Epoch 501/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 410.6217 - mean_absolute_error: 12.3140 - mean_absolute_percentage_error: 13.6148\n", "Epoch 502/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 410.4261 - mean_absolute_error: 12.3408 - mean_absolute_percentage_error: 13.6101\n", "Epoch 503/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.8882 - mean_absolute_error: 12.2583 - mean_absolute_percentage_error: 13.5307\n", "Epoch 504/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 408.2384 - mean_absolute_error: 12.2591 - mean_absolute_percentage_error: 13.5478\n", "Epoch 505/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 407.8354 - mean_absolute_error: 12.2533 - mean_absolute_percentage_error: 13.5117\n", "Epoch 506/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 408.7199 - mean_absolute_error: 12.2684 - mean_absolute_percentage_error: 13.5234\n", "Epoch 507/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.5184 - mean_absolute_error: 12.3403 - mean_absolute_percentage_error: 13.6546\n", "Epoch 508/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.5937 - mean_absolute_error: 12.2759 - mean_absolute_percentage_error: 13.5536\n", "Epoch 509/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 408.5408 - mean_absolute_error: 12.2294 - mean_absolute_percentage_error: 13.5132\n", "Epoch 510/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 407.0204 - mean_absolute_error: 12.2100 - mean_absolute_percentage_error: 13.4906\n", "Epoch 511/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.2841 - mean_absolute_error: 12.2495 - mean_absolute_percentage_error: 13.5418\n", "Epoch 512/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 408.7267 - mean_absolute_error: 12.2848 - mean_absolute_percentage_error: 13.6033\n", "Epoch 513/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 409.3647 - mean_absolute_error: 12.3390 - mean_absolute_percentage_error: 13.5725\n", "Epoch 514/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 410.2119 - mean_absolute_error: 12.2884 - mean_absolute_percentage_error: 13.5424\n", "Epoch 515/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 411.2092 - mean_absolute_error: 12.3353 - mean_absolute_percentage_error: 13.5608\n", "Epoch 516/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 406.5464 - mean_absolute_error: 12.2432 - mean_absolute_percentage_error: 13.5443\n", "Epoch 517/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 407.3925 - mean_absolute_error: 12.2859 - mean_absolute_percentage_error: 13.5984\n", "Epoch 518/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 405.9521 - mean_absolute_error: 12.2339 - mean_absolute_percentage_error: 13.5365\n", "Epoch 519/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 406.8503 - mean_absolute_error: 12.2226 - mean_absolute_percentage_error: 13.4741\n", "Epoch 520/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 406.4733 - mean_absolute_error: 12.1829 - mean_absolute_percentage_error: 13.4667\n", "Epoch 521/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 406.1082 - mean_absolute_error: 12.1822 - mean_absolute_percentage_error: 13.4912\n", "Epoch 522/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 406.3135 - mean_absolute_error: 12.2273 - mean_absolute_percentage_error: 13.5129\n", "Epoch 523/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 406.5419 - mean_absolute_error: 12.2172 - mean_absolute_percentage_error: 13.5323\n", "Epoch 524/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 408.6250 - mean_absolute_error: 12.3629 - mean_absolute_percentage_error: 13.6637\n", "Epoch 525/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 405.8088 - mean_absolute_error: 12.2256 - mean_absolute_percentage_error: 13.5512\n", "Epoch 526/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 405.9876 - mean_absolute_error: 12.2550 - mean_absolute_percentage_error: 13.5973\n", "Epoch 527/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 405.0533 - mean_absolute_error: 12.1993 - mean_absolute_percentage_error: 13.4887\n", "Epoch 528/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 405.6215 - mean_absolute_error: 12.1654 - mean_absolute_percentage_error: 13.4266\n", "Epoch 529/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 405.0737 - mean_absolute_error: 12.2093 - mean_absolute_percentage_error: 13.4812\n", "Epoch 530/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 407.2393 - mean_absolute_error: 12.3232 - mean_absolute_percentage_error: 13.6002\n", "Epoch 531/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.8344 - mean_absolute_error: 12.2039 - mean_absolute_percentage_error: 13.5325\n", "Epoch 532/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 405.5307 - mean_absolute_error: 12.2205 - mean_absolute_percentage_error: 13.5109\n", "Epoch 533/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 405.0576 - mean_absolute_error: 12.2266 - mean_absolute_percentage_error: 13.5299\n", "Epoch 534/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.1055 - mean_absolute_error: 12.2239 - mean_absolute_percentage_error: 13.5412\n", "Epoch 535/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.9548 - mean_absolute_error: 12.2248 - mean_absolute_percentage_error: 13.5642\n", "Epoch 536/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.7211 - mean_absolute_error: 12.2331 - mean_absolute_percentage_error: 13.5479\n", "Epoch 537/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.4004 - mean_absolute_error: 12.2236 - mean_absolute_percentage_error: 13.5378\n", "Epoch 538/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.5067 - mean_absolute_error: 12.1796 - mean_absolute_percentage_error: 13.5063\n", "Epoch 539/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 403.8443 - mean_absolute_error: 12.1599 - mean_absolute_percentage_error: 13.4675\n", "Epoch 540/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.0718 - mean_absolute_error: 12.2169 - mean_absolute_percentage_error: 13.5288\n", "Epoch 541/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 405.2207 - mean_absolute_error: 12.2390 - mean_absolute_percentage_error: 13.5521\n", "Epoch 542/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.6978 - mean_absolute_error: 12.2227 - mean_absolute_percentage_error: 13.5385\n", "Epoch 543/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.7695 - mean_absolute_error: 12.2182 - mean_absolute_percentage_error: 13.5338\n", "Epoch 544/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.5038 - mean_absolute_error: 12.2329 - mean_absolute_percentage_error: 13.5215\n", "Epoch 545/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.5922 - mean_absolute_error: 12.2711 - mean_absolute_percentage_error: 13.6445\n", "Epoch 546/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.7044 - mean_absolute_error: 12.2078 - mean_absolute_percentage_error: 13.5457\n", "Epoch 547/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.3019 - mean_absolute_error: 12.2394 - mean_absolute_percentage_error: 13.5747\n", "Epoch 548/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.6228 - mean_absolute_error: 12.2455 - mean_absolute_percentage_error: 13.5692\n", "Epoch 549/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.0801 - mean_absolute_error: 12.2060 - mean_absolute_percentage_error: 13.5505\n", "Epoch 550/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 402.1650 - mean_absolute_error: 12.1849 - mean_absolute_percentage_error: 13.5259\n", "Epoch 551/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 402.9072 - mean_absolute_error: 12.2115 - mean_absolute_percentage_error: 13.5197\n", "Epoch 552/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.0453 - mean_absolute_error: 12.2297 - mean_absolute_percentage_error: 13.5529\n", "Epoch 553/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.4321 - mean_absolute_error: 12.2500 - mean_absolute_percentage_error: 13.6095\n", "Epoch 554/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 402.9380 - mean_absolute_error: 12.2503 - mean_absolute_percentage_error: 13.5600\n", "Epoch 555/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.4235 - mean_absolute_error: 12.2497 - mean_absolute_percentage_error: 13.5559\n", "Epoch 556/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 402.7479 - mean_absolute_error: 12.2184 - mean_absolute_percentage_error: 13.5541\n", "Epoch 557/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 401.6769 - mean_absolute_error: 12.2092 - mean_absolute_percentage_error: 13.5899\n", "Epoch 558/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 404.7743 - mean_absolute_error: 12.2788 - mean_absolute_percentage_error: 13.5645\n", "Epoch 559/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 403.4757 - mean_absolute_error: 12.1986 - mean_absolute_percentage_error: 13.5188\n", "Epoch 560/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 402.4429 - mean_absolute_error: 12.2353 - mean_absolute_percentage_error: 13.6004\n", "Epoch 561/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 402.0119 - mean_absolute_error: 12.2440 - mean_absolute_percentage_error: 13.6056\n", "Epoch 562/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 401.4492 - mean_absolute_error: 12.2093 - mean_absolute_percentage_error: 13.5437\n", "Epoch 563/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 401.7581 - mean_absolute_error: 12.2428 - mean_absolute_percentage_error: 13.5938\n", "Epoch 564/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 400.9651 - mean_absolute_error: 12.1938 - mean_absolute_percentage_error: 13.5598\n", "Epoch 565/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 402.0729 - mean_absolute_error: 12.2573 - mean_absolute_percentage_error: 13.5917\n", "Epoch 566/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 401.1247 - mean_absolute_error: 12.1963 - mean_absolute_percentage_error: 13.5266\n", "Epoch 567/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 401.9046 - mean_absolute_error: 12.2095 - mean_absolute_percentage_error: 13.6006\n", "Epoch 568/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 402.1708 - mean_absolute_error: 12.2817 - mean_absolute_percentage_error: 13.5901\n", "Epoch 569/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 402.3589 - mean_absolute_error: 12.2592 - mean_absolute_percentage_error: 13.6026\n", "Epoch 570/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 400.0122 - mean_absolute_error: 12.1982 - mean_absolute_percentage_error: 13.5547\n", "Epoch 571/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 402.6152 - mean_absolute_error: 12.3100 - mean_absolute_percentage_error: 13.7102\n", "Epoch 572/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 400.5829 - mean_absolute_error: 12.2245 - mean_absolute_percentage_error: 13.5497\n", "Epoch 573/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 400.4727 - mean_absolute_error: 12.1767 - mean_absolute_percentage_error: 13.5152\n", "Epoch 574/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 400.3094 - mean_absolute_error: 12.1907 - mean_absolute_percentage_error: 13.5724\n", "Epoch 575/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 401.0760 - mean_absolute_error: 12.2756 - mean_absolute_percentage_error: 13.6404\n", "Epoch 576/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 400.1194 - mean_absolute_error: 12.2529 - mean_absolute_percentage_error: 13.6495\n", "Epoch 577/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 400.9396 - mean_absolute_error: 12.2348 - mean_absolute_percentage_error: 13.5903\n", "Epoch 578/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 400.4345 - mean_absolute_error: 12.2104 - mean_absolute_percentage_error: 13.5690\n", "Epoch 579/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 399.8029 - mean_absolute_error: 12.2364 - mean_absolute_percentage_error: 13.6009\n", "Epoch 580/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 399.5345 - mean_absolute_error: 12.2173 - mean_absolute_percentage_error: 13.5844\n", "Epoch 581/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 399.4802 - mean_absolute_error: 12.2190 - mean_absolute_percentage_error: 13.5578\n", "Epoch 582/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 399.7145 - mean_absolute_error: 12.2585 - mean_absolute_percentage_error: 13.5987\n", "Epoch 583/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 398.9418 - mean_absolute_error: 12.1893 - mean_absolute_percentage_error: 13.5336\n", "Epoch 584/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 398.1797 - mean_absolute_error: 12.2294 - mean_absolute_percentage_error: 13.5831\n", "Epoch 585/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 398.5034 - mean_absolute_error: 12.2615 - mean_absolute_percentage_error: 13.6583\n", "Epoch 586/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 399.7737 - mean_absolute_error: 12.2265 - mean_absolute_percentage_error: 13.5657\n", "Epoch 587/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 399.1883 - mean_absolute_error: 12.2381 - mean_absolute_percentage_error: 13.6304\n", "Epoch 588/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 399.5116 - mean_absolute_error: 12.2854 - mean_absolute_percentage_error: 13.6447\n", "Epoch 589/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 397.8005 - mean_absolute_error: 12.2007 - mean_absolute_percentage_error: 13.5759\n", "Epoch 590/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.9283 - mean_absolute_error: 12.1950 - mean_absolute_percentage_error: 13.5485\n", "Epoch 591/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 398.4044 - mean_absolute_error: 12.2219 - mean_absolute_percentage_error: 13.5730\n", "Epoch 592/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 397.8875 - mean_absolute_error: 12.2515 - mean_absolute_percentage_error: 13.6545\n", "Epoch 593/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 398.2068 - mean_absolute_error: 12.2404 - mean_absolute_percentage_error: 13.6033\n", "Epoch 594/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 398.5807 - mean_absolute_error: 12.2183 - mean_absolute_percentage_error: 13.5604\n", "Epoch 595/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 397.1749 - mean_absolute_error: 12.1997 - mean_absolute_percentage_error: 13.5663\n", "Epoch 596/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 395.8564 - mean_absolute_error: 12.2068 - mean_absolute_percentage_error: 13.5855\n", "Epoch 597/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 396.2026 - mean_absolute_error: 12.1990 - mean_absolute_percentage_error: 13.5824\n", "Epoch 598/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 397.3206 - mean_absolute_error: 12.2096 - mean_absolute_percentage_error: 13.6042\n", "Epoch 599/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.7267 - mean_absolute_error: 12.2197 - mean_absolute_percentage_error: 13.6331\n", "Epoch 600/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.6363 - mean_absolute_error: 12.1941 - mean_absolute_percentage_error: 13.5495\n", "Epoch 601/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 395.8442 - mean_absolute_error: 12.1784 - mean_absolute_percentage_error: 13.5632\n", "Epoch 602/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 396.4107 - mean_absolute_error: 12.1798 - mean_absolute_percentage_error: 13.5332\n", "Epoch 603/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 397.2043 - mean_absolute_error: 12.2079 - mean_absolute_percentage_error: 13.5768\n", "Epoch 604/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 395.5493 - mean_absolute_error: 12.1769 - mean_absolute_percentage_error: 13.5834\n", "Epoch 605/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.4536 - mean_absolute_error: 12.1565 - mean_absolute_percentage_error: 13.5493\n", "Epoch 606/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 394.9161 - mean_absolute_error: 12.1669 - mean_absolute_percentage_error: 13.5574\n", "Epoch 607/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.1512 - mean_absolute_error: 12.2040 - mean_absolute_percentage_error: 13.5946\n", "Epoch 608/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 397.6978 - mean_absolute_error: 12.2674 - mean_absolute_percentage_error: 13.6030\n", "Epoch 609/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.8925 - mean_absolute_error: 12.2344 - mean_absolute_percentage_error: 13.6234\n", "Epoch 610/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 400.5656 - mean_absolute_error: 12.3104 - mean_absolute_percentage_error: 13.6318\n", "Epoch 611/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.9084 - mean_absolute_error: 12.2431 - mean_absolute_percentage_error: 13.6359\n", "Epoch 612/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.9070 - mean_absolute_error: 12.2290 - mean_absolute_percentage_error: 13.5947\n", "Epoch 613/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.9481 - mean_absolute_error: 12.2343 - mean_absolute_percentage_error: 13.5764\n", "Epoch 614/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 395.4949 - mean_absolute_error: 12.2057 - mean_absolute_percentage_error: 13.6044\n", "Epoch 615/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 394.9589 - mean_absolute_error: 12.1901 - mean_absolute_percentage_error: 13.5807\n", "Epoch 616/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 394.6515 - mean_absolute_error: 12.1673 - mean_absolute_percentage_error: 13.5653\n", "Epoch 617/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 395.0965 - mean_absolute_error: 12.1761 - mean_absolute_percentage_error: 13.5893\n", "Epoch 618/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 395.6177 - mean_absolute_error: 12.1660 - mean_absolute_percentage_error: 13.5286\n", "Epoch 619/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 395.2088 - mean_absolute_error: 12.2449 - mean_absolute_percentage_error: 13.6744\n", "Epoch 620/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.0524 - mean_absolute_error: 12.2254 - mean_absolute_percentage_error: 13.5784\n", "Epoch 621/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 394.8979 - mean_absolute_error: 12.1488 - mean_absolute_percentage_error: 13.5297\n", "Epoch 622/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 394.2877 - mean_absolute_error: 12.1580 - mean_absolute_percentage_error: 13.5517\n", "Epoch 623/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 393.5189 - mean_absolute_error: 12.1480 - mean_absolute_percentage_error: 13.5278\n", "Epoch 624/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 393.6901 - mean_absolute_error: 12.1766 - mean_absolute_percentage_error: 13.5626\n", "Epoch 625/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 394.3471 - mean_absolute_error: 12.2007 - mean_absolute_percentage_error: 13.5950\n", "Epoch 626/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 396.0023 - mean_absolute_error: 12.1876 - mean_absolute_percentage_error: 13.5823\n", "Epoch 627/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 393.5184 - mean_absolute_error: 12.1793 - mean_absolute_percentage_error: 13.5668\n", "Epoch 628/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 393.5211 - mean_absolute_error: 12.1425 - mean_absolute_percentage_error: 13.5477\n", "Epoch 629/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 393.8872 - mean_absolute_error: 12.1549 - mean_absolute_percentage_error: 13.5930\n", "Epoch 630/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 392.9961 - mean_absolute_error: 12.1384 - mean_absolute_percentage_error: 13.5093\n", "Epoch 631/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 394.0455 - mean_absolute_error: 12.1465 - mean_absolute_percentage_error: 13.5340\n", "Epoch 632/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 393.6738 - mean_absolute_error: 12.1372 - mean_absolute_percentage_error: 13.4987\n", "Epoch 633/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.8326 - mean_absolute_error: 12.1268 - mean_absolute_percentage_error: 13.5201\n", "Epoch 634/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 394.5186 - mean_absolute_error: 12.1718 - mean_absolute_percentage_error: 13.5754\n", "Epoch 635/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 394.0078 - mean_absolute_error: 12.1953 - mean_absolute_percentage_error: 13.5632\n", "Epoch 636/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 393.4293 - mean_absolute_error: 12.1845 - mean_absolute_percentage_error: 13.5749\n", "Epoch 637/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 393.1235 - mean_absolute_error: 12.1525 - mean_absolute_percentage_error: 13.5374\n", "Epoch 638/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.5277 - mean_absolute_error: 12.1436 - mean_absolute_percentage_error: 13.5514\n", "Epoch 639/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.8428 - mean_absolute_error: 12.1427 - mean_absolute_percentage_error: 13.5294\n", "Epoch 640/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.6425 - mean_absolute_error: 12.1457 - mean_absolute_percentage_error: 13.5419\n", "Epoch 641/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.6488 - mean_absolute_error: 12.1553 - mean_absolute_percentage_error: 13.5394\n", "Epoch 642/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 395.9870 - mean_absolute_error: 12.2243 - mean_absolute_percentage_error: 13.5936\n", "Epoch 643/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.7944 - mean_absolute_error: 12.1276 - mean_absolute_percentage_error: 13.5272\n", "Epoch 644/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.8028 - mean_absolute_error: 12.1731 - mean_absolute_percentage_error: 13.5425\n", "Epoch 645/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.7817 - mean_absolute_error: 12.1473 - mean_absolute_percentage_error: 13.5474\n", "Epoch 646/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.0309 - mean_absolute_error: 12.1407 - mean_absolute_percentage_error: 13.5272\n", "Epoch 647/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.1463 - mean_absolute_error: 12.1538 - mean_absolute_percentage_error: 13.5427\n", "Epoch 648/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.9204 - mean_absolute_error: 12.1199 - mean_absolute_percentage_error: 13.5183\n", "Epoch 649/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 391.1079 - mean_absolute_error: 12.0872 - mean_absolute_percentage_error: 13.4831\n", "Epoch 650/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 392.2447 - mean_absolute_error: 12.1435 - mean_absolute_percentage_error: 13.4925\n", "Epoch 651/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 391.9121 - mean_absolute_error: 12.1530 - mean_absolute_percentage_error: 13.5343\n", "Epoch 652/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 393.7421 - mean_absolute_error: 12.1991 - mean_absolute_percentage_error: 13.5400\n", "Epoch 653/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.5162 - mean_absolute_error: 12.1710 - mean_absolute_percentage_error: 13.5836\n", "Epoch 654/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.1020 - mean_absolute_error: 12.1434 - mean_absolute_percentage_error: 13.5289\n", "Epoch 655/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.5794 - mean_absolute_error: 12.1123 - mean_absolute_percentage_error: 13.4888\n", "Epoch 656/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.5215 - mean_absolute_error: 12.1848 - mean_absolute_percentage_error: 13.5771\n", "Epoch 657/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.4954 - mean_absolute_error: 12.0953 - mean_absolute_percentage_error: 13.4855\n", "Epoch 658/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.2912 - mean_absolute_error: 12.1134 - mean_absolute_percentage_error: 13.4913\n", "Epoch 659/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.2557 - mean_absolute_error: 12.1068 - mean_absolute_percentage_error: 13.4760\n", "Epoch 660/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.9718 - mean_absolute_error: 12.1232 - mean_absolute_percentage_error: 13.5375\n", "Epoch 661/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.4128 - mean_absolute_error: 12.1321 - mean_absolute_percentage_error: 13.5447\n", "Epoch 662/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.1064 - mean_absolute_error: 12.1347 - mean_absolute_percentage_error: 13.5239\n", "Epoch 663/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.0535 - mean_absolute_error: 12.0835 - mean_absolute_percentage_error: 13.4807\n", "Epoch 664/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.4826 - mean_absolute_error: 12.1649 - mean_absolute_percentage_error: 13.5434\n", "Epoch 665/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.3833 - mean_absolute_error: 12.1214 - mean_absolute_percentage_error: 13.5232\n", "Epoch 666/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 392.3853 - mean_absolute_error: 12.1872 - mean_absolute_percentage_error: 13.5437\n", "Epoch 667/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.6110 - mean_absolute_error: 12.1332 - mean_absolute_percentage_error: 13.5502\n", "Epoch 668/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.2093 - mean_absolute_error: 12.0841 - mean_absolute_percentage_error: 13.4709\n", "Epoch 669/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.2343 - mean_absolute_error: 12.1026 - mean_absolute_percentage_error: 13.5032\n", "Epoch 670/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.5711 - mean_absolute_error: 12.1202 - mean_absolute_percentage_error: 13.5314\n", "Epoch 671/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.0708 - mean_absolute_error: 12.1058 - mean_absolute_percentage_error: 13.5175\n", "Epoch 672/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 389.5631 - mean_absolute_error: 12.0827 - mean_absolute_percentage_error: 13.4713\n", "Epoch 673/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.2530 - mean_absolute_error: 12.0599 - mean_absolute_percentage_error: 13.4301\n", "Epoch 674/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.2943 - mean_absolute_error: 12.1177 - mean_absolute_percentage_error: 13.4870\n", "Epoch 675/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.7426 - mean_absolute_error: 12.1564 - mean_absolute_percentage_error: 13.5772\n", "Epoch 676/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.1333 - mean_absolute_error: 12.1810 - mean_absolute_percentage_error: 13.5223\n", "Epoch 677/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 393.3734 - mean_absolute_error: 12.2003 - mean_absolute_percentage_error: 13.5683\n", "Epoch 678/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 388.9533 - mean_absolute_error: 12.0886 - mean_absolute_percentage_error: 13.4880\n", "Epoch 679/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.9035 - mean_absolute_error: 12.1998 - mean_absolute_percentage_error: 13.5585\n", "Epoch 680/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 390.9473 - mean_absolute_error: 12.1500 - mean_absolute_percentage_error: 13.5341\n", "Epoch 681/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 390.3493 - mean_absolute_error: 12.1133 - mean_absolute_percentage_error: 13.5375\n", "Epoch 682/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.0679 - mean_absolute_error: 12.1245 - mean_absolute_percentage_error: 13.5246\n", "Epoch 683/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 391.9610 - mean_absolute_error: 12.1887 - mean_absolute_percentage_error: 13.4946\n", "Epoch 684/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.3449 - mean_absolute_error: 12.1882 - mean_absolute_percentage_error: 13.5940\n", "Epoch 685/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 390.3940 - mean_absolute_error: 12.1235 - mean_absolute_percentage_error: 13.5152\n", "Epoch 686/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 389.1913 - mean_absolute_error: 12.1234 - mean_absolute_percentage_error: 13.5234\n", "Epoch 687/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 389.4318 - mean_absolute_error: 12.1149 - mean_absolute_percentage_error: 13.4960\n", "Epoch 688/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 388.9824 - mean_absolute_error: 12.0875 - mean_absolute_percentage_error: 13.4808\n", "Epoch 689/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 388.8149 - mean_absolute_error: 12.1174 - mean_absolute_percentage_error: 13.5343\n", "Epoch 690/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 389.2653 - mean_absolute_error: 12.1199 - mean_absolute_percentage_error: 13.4973\n", "Epoch 691/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 388.6862 - mean_absolute_error: 12.0626 - mean_absolute_percentage_error: 13.4523\n", "Epoch 692/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 389.3043 - mean_absolute_error: 12.1188 - mean_absolute_percentage_error: 13.5436\n", "Epoch 693/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 389.3422 - mean_absolute_error: 12.1355 - mean_absolute_percentage_error: 13.5356\n", "Epoch 694/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 388.2254 - mean_absolute_error: 12.0757 - mean_absolute_percentage_error: 13.4533\n", "Epoch 695/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 388.3299 - mean_absolute_error: 12.0586 - mean_absolute_percentage_error: 13.4558\n", "Epoch 696/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.8034 - mean_absolute_error: 12.0570 - mean_absolute_percentage_error: 13.4951\n", "Epoch 697/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.8216 - mean_absolute_error: 12.0887 - mean_absolute_percentage_error: 13.5154\n", "Epoch 698/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 388.5335 - mean_absolute_error: 12.1382 - mean_absolute_percentage_error: 13.5482\n", "Epoch 699/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 388.0928 - mean_absolute_error: 12.0672 - mean_absolute_percentage_error: 13.4322\n", "Epoch 700/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.9705 - mean_absolute_error: 12.0522 - mean_absolute_percentage_error: 13.4702\n", "Epoch 701/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.6583 - mean_absolute_error: 12.0531 - mean_absolute_percentage_error: 13.4298\n", "Epoch 702/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 388.2109 - mean_absolute_error: 12.1168 - mean_absolute_percentage_error: 13.5343\n", "Epoch 703/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.9914 - mean_absolute_error: 12.0997 - mean_absolute_percentage_error: 13.5086\n", "Epoch 704/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.2646 - mean_absolute_error: 12.0803 - mean_absolute_percentage_error: 13.5123\n", "Epoch 705/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.6698 - mean_absolute_error: 12.0646 - mean_absolute_percentage_error: 13.5041\n", "Epoch 706/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 387.7910 - mean_absolute_error: 12.0835 - mean_absolute_percentage_error: 13.4824\n", "Epoch 707/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 387.2912 - mean_absolute_error: 12.0983 - mean_absolute_percentage_error: 13.5122\n", "Epoch 708/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 386.4885 - mean_absolute_error: 12.0626 - mean_absolute_percentage_error: 13.4857\n", "Epoch 709/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.4936 - mean_absolute_error: 12.0846 - mean_absolute_percentage_error: 13.4782\n", "Epoch 710/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 386.8546 - mean_absolute_error: 12.0710 - mean_absolute_percentage_error: 13.4828\n", "Epoch 711/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.8160 - mean_absolute_error: 12.1158 - mean_absolute_percentage_error: 13.5307\n", "Epoch 712/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.2439 - mean_absolute_error: 12.0403 - mean_absolute_percentage_error: 13.4595\n", "Epoch 713/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 387.1702 - mean_absolute_error: 12.0988 - mean_absolute_percentage_error: 13.5048\n", "Epoch 714/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.5078 - mean_absolute_error: 12.1165 - mean_absolute_percentage_error: 13.5384\n", "Epoch 715/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 386.9274 - mean_absolute_error: 12.0807 - mean_absolute_percentage_error: 13.4759\n", "Epoch 716/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 388.1219 - mean_absolute_error: 12.1243 - mean_absolute_percentage_error: 13.5494\n", "Epoch 717/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 385.8269 - mean_absolute_error: 12.0213 - mean_absolute_percentage_error: 13.4456\n", "Epoch 718/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.8668 - mean_absolute_error: 12.1123 - mean_absolute_percentage_error: 13.5038\n", "Epoch 719/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 387.2492 - mean_absolute_error: 12.0793 - mean_absolute_percentage_error: 13.4902\n", "Epoch 720/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 386.7639 - mean_absolute_error: 12.0421 - mean_absolute_percentage_error: 13.4217\n", "Epoch 721/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 385.6190 - mean_absolute_error: 12.0628 - mean_absolute_percentage_error: 13.4758\n", "Epoch 722/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 386.2304 - mean_absolute_error: 12.1007 - mean_absolute_percentage_error: 13.5558\n", "Epoch 723/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 386.1784 - mean_absolute_error: 12.1143 - mean_absolute_percentage_error: 13.5387\n", "Epoch 724/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 385.5595 - mean_absolute_error: 12.0415 - mean_absolute_percentage_error: 13.4537\n", "Epoch 725/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 386.3880 - mean_absolute_error: 12.0457 - mean_absolute_percentage_error: 13.4351\n", "Epoch 726/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.1555 - mean_absolute_error: 12.0922 - mean_absolute_percentage_error: 13.4947\n", "Epoch 727/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 387.5297 - mean_absolute_error: 12.1452 - mean_absolute_percentage_error: 13.5640\n", "Epoch 728/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 385.3599 - mean_absolute_error: 12.0810 - mean_absolute_percentage_error: 13.5081\n", "Epoch 729/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 386.1184 - mean_absolute_error: 12.0725 - mean_absolute_percentage_error: 13.5143\n", "Epoch 730/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 385.4099 - mean_absolute_error: 12.0542 - mean_absolute_percentage_error: 13.4643\n", "Epoch 731/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.8574 - mean_absolute_error: 12.0590 - mean_absolute_percentage_error: 13.4884\n", "Epoch 732/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.6298 - mean_absolute_error: 12.0484 - mean_absolute_percentage_error: 13.5000\n", "Epoch 733/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.5930 - mean_absolute_error: 12.0578 - mean_absolute_percentage_error: 13.4468\n", "Epoch 734/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.1809 - mean_absolute_error: 12.0636 - mean_absolute_percentage_error: 13.4781\n", "Epoch 735/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.0768 - mean_absolute_error: 12.1006 - mean_absolute_percentage_error: 13.5310\n", "Epoch 736/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 385.0086 - mean_absolute_error: 12.1215 - mean_absolute_percentage_error: 13.5679\n", "Epoch 737/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.1812 - mean_absolute_error: 12.1041 - mean_absolute_percentage_error: 13.5415\n", "Epoch 738/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.7130 - mean_absolute_error: 12.1155 - mean_absolute_percentage_error: 13.5528\n", "Epoch 739/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.5633 - mean_absolute_error: 12.0864 - mean_absolute_percentage_error: 13.5140\n", "Epoch 740/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 383.7274 - mean_absolute_error: 12.0954 - mean_absolute_percentage_error: 13.5695\n", "Epoch 741/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.4080 - mean_absolute_error: 12.1041 - mean_absolute_percentage_error: 13.5803\n", "Epoch 742/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 382.7390 - mean_absolute_error: 12.0826 - mean_absolute_percentage_error: 13.5258\n", "Epoch 743/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.6916 - mean_absolute_error: 12.1357 - mean_absolute_percentage_error: 13.5780\n", "Epoch 744/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.2259 - mean_absolute_error: 12.1254 - mean_absolute_percentage_error: 13.6051\n", "Epoch 745/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.1082 - mean_absolute_error: 12.1499 - mean_absolute_percentage_error: 13.5795\n", "Epoch 746/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 382.7890 - mean_absolute_error: 12.1234 - mean_absolute_percentage_error: 13.5962\n", "Epoch 747/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.2758 - mean_absolute_error: 12.1230 - mean_absolute_percentage_error: 13.6381\n", "Epoch 748/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.9976 - mean_absolute_error: 12.0724 - mean_absolute_percentage_error: 13.5725\n", "Epoch 749/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.8784 - mean_absolute_error: 12.0838 - mean_absolute_percentage_error: 13.5487\n", "Epoch 750/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.7950 - mean_absolute_error: 12.0738 - mean_absolute_percentage_error: 13.5567\n", "Epoch 751/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 382.5077 - mean_absolute_error: 12.0913 - mean_absolute_percentage_error: 13.5829\n", "Epoch 752/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 382.6987 - mean_absolute_error: 12.1002 - mean_absolute_percentage_error: 13.5744\n", "Epoch 753/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.8139 - mean_absolute_error: 12.0821 - mean_absolute_percentage_error: 13.5558\n", "Epoch 754/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.1440 - mean_absolute_error: 12.1224 - mean_absolute_percentage_error: 13.5582\n", "Epoch 755/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.9026 - mean_absolute_error: 12.1507 - mean_absolute_percentage_error: 13.6069\n", "Epoch 756/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.6495 - mean_absolute_error: 12.1799 - mean_absolute_percentage_error: 13.6553\n", "Epoch 757/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 384.3031 - mean_absolute_error: 12.1594 - mean_absolute_percentage_error: 13.5816\n", "Epoch 758/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 382.8910 - mean_absolute_error: 12.1016 - mean_absolute_percentage_error: 13.5640\n", "Epoch 759/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.9818 - mean_absolute_error: 12.1147 - mean_absolute_percentage_error: 13.5812\n", "Epoch 760/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.7090 - mean_absolute_error: 12.1005 - mean_absolute_percentage_error: 13.5789\n", "Epoch 761/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 382.2813 - mean_absolute_error: 12.1022 - mean_absolute_percentage_error: 13.6028\n", "Epoch 762/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.5262 - mean_absolute_error: 12.0665 - mean_absolute_percentage_error: 13.5524\n", "Epoch 763/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.5698 - mean_absolute_error: 12.0827 - mean_absolute_percentage_error: 13.5299\n", "Epoch 764/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 382.9879 - mean_absolute_error: 12.1148 - mean_absolute_percentage_error: 13.6024\n", "Epoch 765/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.4495 - mean_absolute_error: 12.1730 - mean_absolute_percentage_error: 13.6329\n", "Epoch 766/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 383.8813 - mean_absolute_error: 12.1902 - mean_absolute_percentage_error: 13.6743\n", "Epoch 767/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 384.9936 - mean_absolute_error: 12.1855 - mean_absolute_percentage_error: 13.5781\n", "Epoch 768/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.5189 - mean_absolute_error: 12.1151 - mean_absolute_percentage_error: 13.5810\n", "Epoch 769/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 383.6458 - mean_absolute_error: 12.2062 - mean_absolute_percentage_error: 13.7156\n", "Epoch 770/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.1535 - mean_absolute_error: 12.0752 - mean_absolute_percentage_error: 13.5689\n", "Epoch 771/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 380.3949 - mean_absolute_error: 12.0511 - mean_absolute_percentage_error: 13.5295\n", "Epoch 772/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 380.8530 - mean_absolute_error: 12.1051 - mean_absolute_percentage_error: 13.6347\n", "Epoch 773/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 380.2621 - mean_absolute_error: 12.0611 - mean_absolute_percentage_error: 13.5934\n", "Epoch 774/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.0103 - mean_absolute_error: 12.0656 - mean_absolute_percentage_error: 13.5305\n", "Epoch 775/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 381.1475 - mean_absolute_error: 12.0730 - mean_absolute_percentage_error: 13.5562\n", "Epoch 776/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.7444 - mean_absolute_error: 12.0740 - mean_absolute_percentage_error: 13.5704\n", "Epoch 777/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 380.9541 - mean_absolute_error: 12.0850 - mean_absolute_percentage_error: 13.6121\n", "Epoch 778/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.9240 - mean_absolute_error: 12.0635 - mean_absolute_percentage_error: 13.5467\n", "Epoch 779/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 380.8385 - mean_absolute_error: 12.0775 - mean_absolute_percentage_error: 13.5667\n", "Epoch 780/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 380.7605 - mean_absolute_error: 12.0821 - mean_absolute_percentage_error: 13.5518\n", "Epoch 781/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.8940 - mean_absolute_error: 12.0691 - mean_absolute_percentage_error: 13.6024\n", "Epoch 782/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.5623 - mean_absolute_error: 12.0571 - mean_absolute_percentage_error: 13.5745\n", "Epoch 783/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 380.0921 - mean_absolute_error: 12.0852 - mean_absolute_percentage_error: 13.5966\n", "Epoch 784/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.6212 - mean_absolute_error: 12.0600 - mean_absolute_percentage_error: 13.5795\n", "Epoch 785/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.2595 - mean_absolute_error: 12.0605 - mean_absolute_percentage_error: 13.5609\n", "Epoch 786/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.4082 - mean_absolute_error: 12.0376 - mean_absolute_percentage_error: 13.5624\n", "Epoch 787/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.2267 - mean_absolute_error: 12.0386 - mean_absolute_percentage_error: 13.5384\n", "Epoch 788/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.2180 - mean_absolute_error: 12.0448 - mean_absolute_percentage_error: 13.5537\n", "Epoch 789/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.7973 - mean_absolute_error: 12.0814 - mean_absolute_percentage_error: 13.5956\n", "Epoch 790/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 380.0070 - mean_absolute_error: 12.1080 - mean_absolute_percentage_error: 13.6093\n", "Epoch 791/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.9769 - mean_absolute_error: 12.0471 - mean_absolute_percentage_error: 13.5495\n", "Epoch 792/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.0259 - mean_absolute_error: 12.0668 - mean_absolute_percentage_error: 13.5681\n", "Epoch 793/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.5751 - mean_absolute_error: 12.0759 - mean_absolute_percentage_error: 13.6212\n", "Epoch 794/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 379.4251 - mean_absolute_error: 12.0723 - mean_absolute_percentage_error: 13.5535\n", "Epoch 795/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 379.4829 - mean_absolute_error: 12.0362 - mean_absolute_percentage_error: 13.5132\n", "Epoch 796/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.5552 - mean_absolute_error: 12.0862 - mean_absolute_percentage_error: 13.6233\n", "Epoch 797/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 381.1110 - mean_absolute_error: 12.1645 - mean_absolute_percentage_error: 13.6825\n", "Epoch 798/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.3577 - mean_absolute_error: 12.0214 - mean_absolute_percentage_error: 13.5545\n", "Epoch 799/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.8385 - mean_absolute_error: 12.0198 - mean_absolute_percentage_error: 13.5126\n", "Epoch 800/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.8844 - mean_absolute_error: 12.0494 - mean_absolute_percentage_error: 13.5637\n", "Epoch 801/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.5432 - mean_absolute_error: 12.0563 - mean_absolute_percentage_error: 13.5992\n", "Epoch 802/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.4790 - mean_absolute_error: 12.0599 - mean_absolute_percentage_error: 13.6078\n", "Epoch 803/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.0434 - mean_absolute_error: 12.0180 - mean_absolute_percentage_error: 13.5375\n", "Epoch 804/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 379.8079 - mean_absolute_error: 12.0447 - mean_absolute_percentage_error: 13.5331\n", "Epoch 805/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.5703 - mean_absolute_error: 12.0404 - mean_absolute_percentage_error: 13.6037\n", "Epoch 806/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.4408 - mean_absolute_error: 12.0519 - mean_absolute_percentage_error: 13.5510\n", "Epoch 807/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.7405 - mean_absolute_error: 12.0402 - mean_absolute_percentage_error: 13.5550\n", "Epoch 808/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.4291 - mean_absolute_error: 12.0618 - mean_absolute_percentage_error: 13.5747\n", "Epoch 809/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.0081 - mean_absolute_error: 12.0624 - mean_absolute_percentage_error: 13.5708\n", "Epoch 810/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.3883 - mean_absolute_error: 12.0425 - mean_absolute_percentage_error: 13.5649\n", "Epoch 811/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.7617 - mean_absolute_error: 12.0063 - mean_absolute_percentage_error: 13.5323\n", "Epoch 812/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.8596 - mean_absolute_error: 12.0370 - mean_absolute_percentage_error: 13.5430\n", "Epoch 813/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.5031 - mean_absolute_error: 12.0768 - mean_absolute_percentage_error: 13.5638\n", "Epoch 814/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.6968 - mean_absolute_error: 12.0536 - mean_absolute_percentage_error: 13.5789\n", "Epoch 815/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.8498 - mean_absolute_error: 12.0395 - mean_absolute_percentage_error: 13.5266\n", "Epoch 816/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 376.7412 - mean_absolute_error: 12.0086 - mean_absolute_percentage_error: 13.5234\n", "Epoch 817/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.6898 - mean_absolute_error: 12.0955 - mean_absolute_percentage_error: 13.6562\n", "Epoch 818/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 376.5739 - mean_absolute_error: 12.0182 - mean_absolute_percentage_error: 13.5624\n", "Epoch 819/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 377.5638 - mean_absolute_error: 12.0333 - mean_absolute_percentage_error: 13.5830\n", "Epoch 820/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 377.4822 - mean_absolute_error: 12.0609 - mean_absolute_percentage_error: 13.5761\n", "Epoch 821/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 378.1814 - mean_absolute_error: 12.0644 - mean_absolute_percentage_error: 13.5623\n", "Epoch 822/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.3946 - mean_absolute_error: 12.0494 - mean_absolute_percentage_error: 13.5841\n", "Epoch 823/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.2269 - mean_absolute_error: 12.0396 - mean_absolute_percentage_error: 13.5402\n", "Epoch 824/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.7704 - mean_absolute_error: 12.0554 - mean_absolute_percentage_error: 13.6028\n", "Epoch 825/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 377.3838 - mean_absolute_error: 12.0698 - mean_absolute_percentage_error: 13.6235\n", "Epoch 826/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 376.6518 - mean_absolute_error: 11.9936 - mean_absolute_percentage_error: 13.5233\n", "Epoch 827/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 376.2243 - mean_absolute_error: 11.9614 - mean_absolute_percentage_error: 13.4887\n", "Epoch 828/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.2543 - mean_absolute_error: 12.0151 - mean_absolute_percentage_error: 13.5323\n", "Epoch 829/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.3535 - mean_absolute_error: 12.0311 - mean_absolute_percentage_error: 13.5664\n", "Epoch 830/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 376.6264 - mean_absolute_error: 12.0527 - mean_absolute_percentage_error: 13.5850\n", "Epoch 831/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 376.7873 - mean_absolute_error: 12.0106 - mean_absolute_percentage_error: 13.5258\n", "Epoch 832/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 376.1347 - mean_absolute_error: 11.9966 - mean_absolute_percentage_error: 13.5238\n", "Epoch 833/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 376.4964 - mean_absolute_error: 12.0146 - mean_absolute_percentage_error: 13.5651\n", "Epoch 834/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 378.2726 - mean_absolute_error: 12.0934 - mean_absolute_percentage_error: 13.5926\n", "Epoch 835/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 379.5754 - mean_absolute_error: 12.1732 - mean_absolute_percentage_error: 13.6474\n", "Epoch 836/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 382.2725 - mean_absolute_error: 12.2151 - mean_absolute_percentage_error: 13.6443\n", "Epoch 837/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.3899 - mean_absolute_error: 12.0693 - mean_absolute_percentage_error: 13.6339\n", "Epoch 838/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 376.5905 - mean_absolute_error: 12.0292 - mean_absolute_percentage_error: 13.5674\n", "Epoch 839/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.7317 - mean_absolute_error: 11.9834 - mean_absolute_percentage_error: 13.5147\n", "Epoch 840/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.1776 - mean_absolute_error: 12.0498 - mean_absolute_percentage_error: 13.5494\n", "Epoch 841/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 377.1432 - mean_absolute_error: 12.0592 - mean_absolute_percentage_error: 13.5915\n", "Epoch 842/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.7260 - mean_absolute_error: 12.0224 - mean_absolute_percentage_error: 13.5796\n", "Epoch 843/1000\n", "13/13 [==============================] - 0s 3ms/step - loss: 376.1706 - mean_absolute_error: 11.9965 - mean_absolute_percentage_error: 13.5804\n", "Epoch 844/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.7959 - mean_absolute_error: 11.9937 - mean_absolute_percentage_error: 13.5452\n", "Epoch 845/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 376.5505 - mean_absolute_error: 11.9885 - mean_absolute_percentage_error: 13.4825\n", "Epoch 846/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.5602 - mean_absolute_error: 11.9952 - mean_absolute_percentage_error: 13.5410\n", "Epoch 847/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.2238 - mean_absolute_error: 12.0158 - mean_absolute_percentage_error: 13.5860\n", "Epoch 848/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.4559 - mean_absolute_error: 11.9713 - mean_absolute_percentage_error: 13.5343\n", "Epoch 849/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.2596 - mean_absolute_error: 12.0023 - mean_absolute_percentage_error: 13.5593\n", "Epoch 850/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.4140 - mean_absolute_error: 11.9726 - mean_absolute_percentage_error: 13.5280\n", "Epoch 851/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.4799 - mean_absolute_error: 11.9826 - mean_absolute_percentage_error: 13.5023\n", "Epoch 852/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.6302 - mean_absolute_error: 11.9877 - mean_absolute_percentage_error: 13.5474\n", "Epoch 853/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.9900 - mean_absolute_error: 12.0204 - mean_absolute_percentage_error: 13.5502\n", "Epoch 854/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.9006 - mean_absolute_error: 11.9706 - mean_absolute_percentage_error: 13.5577\n", "Epoch 855/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.3855 - mean_absolute_error: 11.9966 - mean_absolute_percentage_error: 13.5427\n", "Epoch 856/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.1094 - mean_absolute_error: 11.9803 - mean_absolute_percentage_error: 13.5227\n", "Epoch 857/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.0732 - mean_absolute_error: 12.0080 - mean_absolute_percentage_error: 13.5601\n", "Epoch 858/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.0717 - mean_absolute_error: 11.9606 - mean_absolute_percentage_error: 13.5161\n", "Epoch 859/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.0109 - mean_absolute_error: 11.9596 - mean_absolute_percentage_error: 13.4901\n", "Epoch 860/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.3424 - mean_absolute_error: 11.9944 - mean_absolute_percentage_error: 13.5371\n", "Epoch 861/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 374.7935 - mean_absolute_error: 11.9968 - mean_absolute_percentage_error: 13.5719\n", "Epoch 862/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.2662 - mean_absolute_error: 11.9653 - mean_absolute_percentage_error: 13.5500\n", "Epoch 863/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.6660 - mean_absolute_error: 11.9722 - mean_absolute_percentage_error: 13.5074\n", "Epoch 864/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.1655 - mean_absolute_error: 11.9625 - mean_absolute_percentage_error: 13.5079\n", "Epoch 865/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.8041 - mean_absolute_error: 11.9880 - mean_absolute_percentage_error: 13.5572\n", "Epoch 866/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 374.2770 - mean_absolute_error: 11.9748 - mean_absolute_percentage_error: 13.5451\n", "Epoch 867/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.7505 - mean_absolute_error: 11.9927 - mean_absolute_percentage_error: 13.5348\n", "Epoch 868/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 375.6982 - mean_absolute_error: 12.0025 - mean_absolute_percentage_error: 13.5259\n", "Epoch 869/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.6196 - mean_absolute_error: 11.9857 - mean_absolute_percentage_error: 13.5637\n", "Epoch 870/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.4521 - mean_absolute_error: 12.0113 - mean_absolute_percentage_error: 13.5893\n", "Epoch 871/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 374.6181 - mean_absolute_error: 11.9881 - mean_absolute_percentage_error: 13.4959\n", "Epoch 872/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.7305 - mean_absolute_error: 11.9461 - mean_absolute_percentage_error: 13.5062\n", "Epoch 873/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 373.7799 - mean_absolute_error: 11.9936 - mean_absolute_percentage_error: 13.5786\n", "Epoch 874/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.6047 - mean_absolute_error: 12.0170 - mean_absolute_percentage_error: 13.6009\n", "Epoch 875/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.0366 - mean_absolute_error: 11.9632 - mean_absolute_percentage_error: 13.4942\n", "Epoch 876/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 374.1517 - mean_absolute_error: 11.9222 - mean_absolute_percentage_error: 13.4904\n", "Epoch 877/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.5794 - mean_absolute_error: 11.9732 - mean_absolute_percentage_error: 13.5796\n", "Epoch 878/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.2655 - mean_absolute_error: 11.9672 - mean_absolute_percentage_error: 13.5182\n", "Epoch 879/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 373.3338 - mean_absolute_error: 11.9330 - mean_absolute_percentage_error: 13.5065\n", "Epoch 880/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.1453 - mean_absolute_error: 12.0085 - mean_absolute_percentage_error: 13.5404\n", "Epoch 881/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.4816 - mean_absolute_error: 12.0108 - mean_absolute_percentage_error: 13.5531\n", "Epoch 882/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.2973 - mean_absolute_error: 11.9448 - mean_absolute_percentage_error: 13.5073\n", "Epoch 883/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.5067 - mean_absolute_error: 11.9870 - mean_absolute_percentage_error: 13.5473\n", "Epoch 884/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.1177 - mean_absolute_error: 11.9619 - mean_absolute_percentage_error: 13.5367\n", "Epoch 885/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 374.3748 - mean_absolute_error: 11.9620 - mean_absolute_percentage_error: 13.5251\n", "Epoch 886/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.8646 - mean_absolute_error: 11.9378 - mean_absolute_percentage_error: 13.5384\n", "Epoch 887/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.9262 - mean_absolute_error: 11.9539 - mean_absolute_percentage_error: 13.5478\n", "Epoch 888/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 372.8244 - mean_absolute_error: 11.9428 - mean_absolute_percentage_error: 13.4953\n", "Epoch 889/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.8729 - mean_absolute_error: 11.9394 - mean_absolute_percentage_error: 13.5072\n", "Epoch 890/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.3562 - mean_absolute_error: 11.9800 - mean_absolute_percentage_error: 13.6009\n", "Epoch 891/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.2422 - mean_absolute_error: 11.9777 - mean_absolute_percentage_error: 13.5817\n", "Epoch 892/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.2738 - mean_absolute_error: 11.9259 - mean_absolute_percentage_error: 13.5030\n", "Epoch 893/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.4875 - mean_absolute_error: 11.9685 - mean_absolute_percentage_error: 13.5334\n", "Epoch 894/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.0145 - mean_absolute_error: 11.9377 - mean_absolute_percentage_error: 13.5108\n", "Epoch 895/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.6097 - mean_absolute_error: 11.9593 - mean_absolute_percentage_error: 13.5569\n", "Epoch 896/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.9030 - mean_absolute_error: 11.9452 - mean_absolute_percentage_error: 13.5623\n", "Epoch 897/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.6751 - mean_absolute_error: 11.9127 - mean_absolute_percentage_error: 13.4884\n", "Epoch 898/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.8644 - mean_absolute_error: 11.9391 - mean_absolute_percentage_error: 13.5050\n", "Epoch 899/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.0750 - mean_absolute_error: 11.9293 - mean_absolute_percentage_error: 13.5282\n", "Epoch 900/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.1368 - mean_absolute_error: 11.9239 - mean_absolute_percentage_error: 13.5364\n", "Epoch 901/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.1086 - mean_absolute_error: 11.9726 - mean_absolute_percentage_error: 13.5329\n", "Epoch 902/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.8742 - mean_absolute_error: 11.9645 - mean_absolute_percentage_error: 13.5583\n", "Epoch 903/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.5091 - mean_absolute_error: 11.9252 - mean_absolute_percentage_error: 13.5559\n", "Epoch 904/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.7176 - mean_absolute_error: 11.9325 - mean_absolute_percentage_error: 13.5036\n", "Epoch 905/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 373.1253 - mean_absolute_error: 11.9767 - mean_absolute_percentage_error: 13.5390\n", "Epoch 906/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.3779 - mean_absolute_error: 11.9740 - mean_absolute_percentage_error: 13.5337\n", "Epoch 907/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.0161 - mean_absolute_error: 11.9368 - mean_absolute_percentage_error: 13.5125\n", "Epoch 908/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.6537 - mean_absolute_error: 11.9726 - mean_absolute_percentage_error: 13.5533\n", "Epoch 909/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.7957 - mean_absolute_error: 11.9352 - mean_absolute_percentage_error: 13.5335\n", "Epoch 910/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.9754 - mean_absolute_error: 11.9510 - mean_absolute_percentage_error: 13.5702\n", "Epoch 911/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.3662 - mean_absolute_error: 11.9399 - mean_absolute_percentage_error: 13.5381\n", "Epoch 912/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.7685 - mean_absolute_error: 11.9449 - mean_absolute_percentage_error: 13.5060\n", "Epoch 913/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.8359 - mean_absolute_error: 11.9684 - mean_absolute_percentage_error: 13.5646\n", "Epoch 914/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.4358 - mean_absolute_error: 11.9707 - mean_absolute_percentage_error: 13.5962\n", "Epoch 915/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 370.9055 - mean_absolute_error: 11.9166 - mean_absolute_percentage_error: 13.5276\n", "Epoch 916/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 370.9901 - mean_absolute_error: 11.9241 - mean_absolute_percentage_error: 13.5332\n", "Epoch 917/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.5198 - mean_absolute_error: 11.9019 - mean_absolute_percentage_error: 13.4739\n", "Epoch 918/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.7391 - mean_absolute_error: 11.9868 - mean_absolute_percentage_error: 13.5752\n", "Epoch 919/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.4236 - mean_absolute_error: 11.9449 - mean_absolute_percentage_error: 13.5667\n", "Epoch 920/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.0643 - mean_absolute_error: 11.9575 - mean_absolute_percentage_error: 13.5335\n", "Epoch 921/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.8588 - mean_absolute_error: 12.0114 - mean_absolute_percentage_error: 13.5946\n", "Epoch 922/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 374.9572 - mean_absolute_error: 12.1070 - mean_absolute_percentage_error: 13.6143\n", "Epoch 923/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.4446 - mean_absolute_error: 12.0163 - mean_absolute_percentage_error: 13.5807\n", "Epoch 924/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.0409 - mean_absolute_error: 11.9794 - mean_absolute_percentage_error: 13.5599\n", "Epoch 925/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 370.1891 - mean_absolute_error: 11.9278 - mean_absolute_percentage_error: 13.5574\n", "Epoch 926/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.5665 - mean_absolute_error: 11.9704 - mean_absolute_percentage_error: 13.5437\n", "Epoch 927/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.5806 - mean_absolute_error: 11.9603 - mean_absolute_percentage_error: 13.5407\n", "Epoch 928/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 371.5778 - mean_absolute_error: 11.9464 - mean_absolute_percentage_error: 13.5811\n", "Epoch 929/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 372.4156 - mean_absolute_error: 12.0157 - mean_absolute_percentage_error: 13.5924\n", "Epoch 930/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.5471 - mean_absolute_error: 12.0280 - mean_absolute_percentage_error: 13.6027\n", "Epoch 931/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.7994 - mean_absolute_error: 12.0273 - mean_absolute_percentage_error: 13.6297\n", "Epoch 932/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 370.6846 - mean_absolute_error: 11.9549 - mean_absolute_percentage_error: 13.5764\n", "Epoch 933/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 371.2269 - mean_absolute_error: 11.9369 - mean_absolute_percentage_error: 13.5194\n", "Epoch 934/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 369.4266 - mean_absolute_error: 11.9344 - mean_absolute_percentage_error: 13.5412\n", "Epoch 935/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 370.0995 - mean_absolute_error: 11.9510 - mean_absolute_percentage_error: 13.5647\n", "Epoch 936/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 370.8477 - mean_absolute_error: 11.9592 - mean_absolute_percentage_error: 13.6244\n", "Epoch 937/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 370.5511 - mean_absolute_error: 11.9711 - mean_absolute_percentage_error: 13.5860\n", "Epoch 938/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.9824 - mean_absolute_error: 11.9541 - mean_absolute_percentage_error: 13.5887\n", "Epoch 939/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.8416 - mean_absolute_error: 11.9716 - mean_absolute_percentage_error: 13.6373\n", "Epoch 940/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.1157 - mean_absolute_error: 11.9370 - mean_absolute_percentage_error: 13.6174\n", "Epoch 941/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.8869 - mean_absolute_error: 11.9568 - mean_absolute_percentage_error: 13.5877\n", "Epoch 942/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.4073 - mean_absolute_error: 11.9545 - mean_absolute_percentage_error: 13.5715\n", "Epoch 943/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.3298 - mean_absolute_error: 11.9298 - mean_absolute_percentage_error: 13.6112\n", "Epoch 944/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 368.2040 - mean_absolute_error: 11.9683 - mean_absolute_percentage_error: 13.6192\n", "Epoch 945/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.9666 - mean_absolute_error: 11.9593 - mean_absolute_percentage_error: 13.6122\n", "Epoch 946/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 369.2866 - mean_absolute_error: 11.9662 - mean_absolute_percentage_error: 13.6111\n", "Epoch 947/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.4193 - mean_absolute_error: 11.9438 - mean_absolute_percentage_error: 13.5765\n", "Epoch 948/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.8096 - mean_absolute_error: 11.9704 - mean_absolute_percentage_error: 13.6001\n", "Epoch 949/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.2166 - mean_absolute_error: 11.9463 - mean_absolute_percentage_error: 13.6194\n", "Epoch 950/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.4833 - mean_absolute_error: 11.9281 - mean_absolute_percentage_error: 13.5835\n", "Epoch 951/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.9065 - mean_absolute_error: 11.8874 - mean_absolute_percentage_error: 13.5469\n", "Epoch 952/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.9966 - mean_absolute_error: 11.9104 - mean_absolute_percentage_error: 13.5507\n", "Epoch 953/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.2837 - mean_absolute_error: 11.9056 - mean_absolute_percentage_error: 13.5553\n", "Epoch 954/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.1404 - mean_absolute_error: 11.8886 - mean_absolute_percentage_error: 13.5790\n", "Epoch 955/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.8309 - mean_absolute_error: 11.9055 - mean_absolute_percentage_error: 13.5798\n", "Epoch 956/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.2178 - mean_absolute_error: 11.9134 - mean_absolute_percentage_error: 13.5659\n", "Epoch 957/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.2767 - mean_absolute_error: 11.9132 - mean_absolute_percentage_error: 13.5709\n", "Epoch 958/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.2410 - mean_absolute_error: 11.9397 - mean_absolute_percentage_error: 13.5818\n", "Epoch 959/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 366.2132 - mean_absolute_error: 11.9161 - mean_absolute_percentage_error: 13.5954\n", "Epoch 960/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.0746 - mean_absolute_error: 11.9125 - mean_absolute_percentage_error: 13.5685\n", "Epoch 961/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.7951 - mean_absolute_error: 11.9269 - mean_absolute_percentage_error: 13.5698\n", "Epoch 962/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.6908 - mean_absolute_error: 11.9138 - mean_absolute_percentage_error: 13.5738\n", "Epoch 963/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.5147 - mean_absolute_error: 11.9534 - mean_absolute_percentage_error: 13.5581\n", "Epoch 964/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 368.7565 - mean_absolute_error: 11.9937 - mean_absolute_percentage_error: 13.6513\n", "Epoch 965/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.1995 - mean_absolute_error: 11.9809 - mean_absolute_percentage_error: 13.6341\n", "Epoch 966/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.6778 - mean_absolute_error: 11.9268 - mean_absolute_percentage_error: 13.5929\n", "Epoch 967/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.2755 - mean_absolute_error: 11.9000 - mean_absolute_percentage_error: 13.5384\n", "Epoch 968/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 367.5467 - mean_absolute_error: 11.9547 - mean_absolute_percentage_error: 13.6438\n", "Epoch 969/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 365.9991 - mean_absolute_error: 11.9126 - mean_absolute_percentage_error: 13.6156\n", "Epoch 970/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.3637 - mean_absolute_error: 11.8915 - mean_absolute_percentage_error: 13.5467\n", "Epoch 971/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 365.6040 - mean_absolute_error: 11.9136 - mean_absolute_percentage_error: 13.5787\n", "Epoch 972/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 365.8547 - mean_absolute_error: 11.8896 - mean_absolute_percentage_error: 13.5862\n", "Epoch 973/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.3911 - mean_absolute_error: 11.9428 - mean_absolute_percentage_error: 13.5922\n", "Epoch 974/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.1723 - mean_absolute_error: 11.9723 - mean_absolute_percentage_error: 13.6212\n", "Epoch 975/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 365.4499 - mean_absolute_error: 11.9079 - mean_absolute_percentage_error: 13.5707\n", "Epoch 976/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 365.6083 - mean_absolute_error: 11.8665 - mean_absolute_percentage_error: 13.5527\n", "Epoch 977/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.9531 - mean_absolute_error: 11.8721 - mean_absolute_percentage_error: 13.5209\n", "Epoch 978/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.7233 - mean_absolute_error: 11.8579 - mean_absolute_percentage_error: 13.5330\n", "Epoch 979/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.9843 - mean_absolute_error: 11.8754 - mean_absolute_percentage_error: 13.5358\n", "Epoch 980/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.1961 - mean_absolute_error: 11.9057 - mean_absolute_percentage_error: 13.6152\n", "Epoch 981/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.8169 - mean_absolute_error: 11.8899 - mean_absolute_percentage_error: 13.5911\n", "Epoch 982/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 364.9774 - mean_absolute_error: 11.8951 - mean_absolute_percentage_error: 13.5590\n", "Epoch 983/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.1364 - mean_absolute_error: 11.9042 - mean_absolute_percentage_error: 13.5604\n", "Epoch 984/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.7208 - mean_absolute_error: 11.8628 - mean_absolute_percentage_error: 13.5253\n", "Epoch 985/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 365.5602 - mean_absolute_error: 11.9027 - mean_absolute_percentage_error: 13.5373\n", "Epoch 986/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.9245 - mean_absolute_error: 11.8648 - mean_absolute_percentage_error: 13.5301\n", "Epoch 987/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.6560 - mean_absolute_error: 11.8464 - mean_absolute_percentage_error: 13.5246\n", "Epoch 988/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.3084 - mean_absolute_error: 11.8412 - mean_absolute_percentage_error: 13.5104\n", "Epoch 989/1000\n", "13/13 [==============================] - 0s 1ms/step - loss: 363.8415 - mean_absolute_error: 11.8406 - mean_absolute_percentage_error: 13.4769\n", "Epoch 990/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.9170 - mean_absolute_error: 11.8423 - mean_absolute_percentage_error: 13.4878\n", "Epoch 991/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.6495 - mean_absolute_error: 11.8603 - mean_absolute_percentage_error: 13.4825\n", "Epoch 992/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.8329 - mean_absolute_error: 11.8361 - mean_absolute_percentage_error: 13.5058\n", "Epoch 993/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.1258 - mean_absolute_error: 11.8757 - mean_absolute_percentage_error: 13.4988\n", "Epoch 994/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.9192 - mean_absolute_error: 11.8533 - mean_absolute_percentage_error: 13.4663\n", "Epoch 995/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.4176 - mean_absolute_error: 11.8313 - mean_absolute_percentage_error: 13.4337\n", "Epoch 996/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 363.9436 - mean_absolute_error: 11.8362 - mean_absolute_percentage_error: 13.4926\n", "Epoch 997/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.1689 - mean_absolute_error: 11.9052 - mean_absolute_percentage_error: 13.5359\n", "Epoch 998/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 364.5861 - mean_absolute_error: 11.9305 - mean_absolute_percentage_error: 13.5271\n", "Epoch 999/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.3006 - mean_absolute_error: 11.9117 - mean_absolute_percentage_error: 13.4541\n", "Epoch 1000/1000\n", "13/13 [==============================] - 0s 2ms/step - loss: 366.4105 - mean_absolute_error: 11.9439 - mean_absolute_percentage_error: 13.5268\n" ], "name": "stdout" } ] }, { "cell_type": "code", "metadata": { "id": "RLUGAjnXtbcB", "outputId": "667da3f3-913f-4510-db39-4ff559b2d2d3", "colab": { "base_uri": "https://localhost:8080/", "height": 54 } }, "source": [ "model.evaluate(x_test_norm, y_test)" ], "execution_count": 18, "outputs": [ { "output_type": "stream", "text": [ "4/4 [==============================] - 0s 3ms/step - loss: 1875.0286 - mean_absolute_error: 27.7769 - mean_absolute_percentage_error: 28.0428\n" ], "name": "stdout" }, { "output_type": "execute_result", "data": { "text/plain": [ "[1875.028564453125, 27.77690315246582, 28.042774200439453]" ] }, "metadata": { "tags": [] }, "execution_count": 18 } ] }, { "cell_type": "code", "metadata": { "id": "CLhZwedGmrCi", "outputId": "9e5e6b04-7578-4b5d-ec64-c5be8c4b2cf1", "colab": { "base_uri": "https://localhost:8080/", "height": 35 } }, "source": [ "history_with_minibatch.history.keys()" ], "execution_count": 19, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "dict_keys(['loss', 'mean_absolute_error', 'mean_absolute_percentage_error'])" ] }, "metadata": { "tags": [] }, "execution_count": 19 } ] }, { "cell_type": "code", "metadata": { "id": "tePZj3exmtJi", "outputId": "de610abf-0758-4d4e-bae3-46bc49b5e0fc", "colab": { "base_uri": "https://localhost:8080/", "height": 328 } }, "source": [ "plt.plot(history_with_minibatch.history['loss'],linewidth=2.0)\n", "plt.title('Training Mean Squared Error (MSE)\\n with architecture (10-20-2) using mini-batch', fontsize=12)\n", "plt.xlabel('epochs')\n", "plt.ylabel('Loss')\n", "#plt.savefig('images/curva', bbox_inches='tight')\n", "#plt.plot(test_history_with_minibatch.losses.T[0])" ], "execution_count": 20, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "Text(0, 0.5, 'Loss')" ] }, "metadata": { "tags": [] }, "execution_count": 20 }, { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [], "needs_background": "light" } } ] }, { "cell_type": "markdown", "metadata": { "id": "bNYOJKRCnMk-" }, "source": [ "The same training is performed using BGD, that means, the size of the batch is the length of the dataset." ] }, { "cell_type": "code", "metadata": { "id": "G_Icyvw8nLGp" }, "source": [ "# Redefining the model not to override the previous results\n", "model2 = models.Sequential()\n", "model2.add(layers.Dense(20, activation='sigmoid', input_shape=(10,)))\n", "model2.add(layers.Dense(2))\n", "\n", "model2.compile(optimizer=sgd,\n", " loss='mean_squared_error',\n", " metrics=['mean_absolute_error', 'mean_absolute_percentage_error'])" ], "execution_count": 25, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "InVA8JFhnQ6J", "outputId": "4e5aaea9-09d0-499c-827a-52343eb73873", "colab": { "base_uri": "https://localhost:8080/", "height": 1000 } }, "source": [ "history_without_minibatch = model2.fit(x_train_norm, y_train, epochs=1000, batch_size=x_train.shape[0])\n", "\n", " \n", "# To use the test loss history, comment the lines above and uncomment the lines below\n", "#test_history_without_minibatch = TestLossHistory(x_test, y_test)\n", "#history_without_minibatch = model2.fit(x_train_norm, y_train, epochs=10000, batch_size=x_train.shape[0],\n", "# callbacks=[test_history_without_minibatch])" ], "execution_count": 26, "outputs": [ { "output_type": "stream", "text": [ "Epoch 1/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 13086.8965 - mean_absolute_error: 96.9786 - mean_absolute_percentage_error: 100.1610\n", "Epoch 2/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 12975.7861 - mean_absolute_error: 96.3969 - mean_absolute_percentage_error: 99.3832\n", "Epoch 3/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 12767.0850 - mean_absolute_error: 95.2946 - mean_absolute_percentage_error: 97.9097\n", "Epoch 4/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 12473.5566 - mean_absolute_error: 93.7237 - mean_absolute_percentage_error: 95.8115\n", "Epoch 5/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 12104.9277 - mean_absolute_error: 91.7183 - mean_absolute_percentage_error: 93.1392\n", "Epoch 6/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 11666.1084 - mean_absolute_error: 89.2865 - mean_absolute_percentage_error: 89.9141\n", "Epoch 7/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 11156.4932 - mean_absolute_error: 86.4054 - mean_absolute_percentage_error: 86.1222\n", "Epoch 8/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 10570.9014 - mean_absolute_error: 83.0200 - mean_absolute_percentage_error: 81.7130\n", "Epoch 9/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 9902.6914 - mean_absolute_error: 79.0497 - mean_absolute_percentage_error: 76.6059\n", "Epoch 10/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 9149.2236 - mean_absolute_error: 74.4050 - mean_absolute_percentage_error: 70.7064\n", "Epoch 11/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 8318.5303 - mean_absolute_error: 69.0139 - mean_absolute_percentage_error: 63.9324\n", "Epoch 12/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 7434.3057 - mean_absolute_error: 62.8523 - mean_absolute_percentage_error: 56.2484\n", "Epoch 13/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 6536.3584 - mean_absolute_error: 55.9703 - mean_absolute_percentage_error: 47.6985\n", "Epoch 14/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 5675.4902 - mean_absolute_error: 48.5653 - mean_absolute_percentage_error: 38.5835\n", "Epoch 15/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 4903.2661 - mean_absolute_error: 41.2661 - mean_absolute_percentage_error: 30.0402\n", "Epoch 16/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 4259.4424 - mean_absolute_error: 36.0362 - mean_absolute_percentage_error: 25.5715\n", "Epoch 17/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3763.2615 - mean_absolute_error: 33.3690 - mean_absolute_percentage_error: 25.1816\n", "Epoch 18/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3412.9397 - mean_absolute_error: 32.8971 - mean_absolute_percentage_error: 27.6768\n", "Epoch 19/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3191.2607 - mean_absolute_error: 33.9621 - mean_absolute_percentage_error: 31.7957\n", "Epoch 20/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3072.5625 - mean_absolute_error: 35.7645 - mean_absolute_percentage_error: 36.4009\n", "Epoch 21/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3028.2827 - mean_absolute_error: 37.6546 - mean_absolute_percentage_error: 40.6669\n", "Epoch 22/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3030.7017 - mean_absolute_error: 39.4898 - mean_absolute_percentage_error: 44.3678\n", "Epoch 23/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 3055.4480 - mean_absolute_error: 41.0316 - mean_absolute_percentage_error: 47.2253\n", "Epoch 24/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3083.1936 - mean_absolute_error: 42.1486 - mean_absolute_percentage_error: 49.1631\n", "Epoch 25/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3100.6016 - mean_absolute_error: 42.8440 - mean_absolute_percentage_error: 50.2409\n", "Epoch 26/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3100.4736 - mean_absolute_error: 43.1868 - mean_absolute_percentage_error: 50.6559\n", "Epoch 27/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 3081.0764 - mean_absolute_error: 43.2213 - mean_absolute_percentage_error: 50.5449\n", "Epoch 28/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 3044.8044 - mean_absolute_error: 42.9041 - mean_absolute_percentage_error: 49.8397\n", "Epoch 29/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2996.4248 - mean_absolute_error: 42.3051 - mean_absolute_percentage_error: 48.7021\n", "Epoch 30/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 2941.3394 - mean_absolute_error: 41.5382 - mean_absolute_percentage_error: 47.3248\n", "Epoch 31/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2884.2847 - mean_absolute_error: 40.7911 - mean_absolute_percentage_error: 46.0643\n", "Epoch 32/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2828.7134 - mean_absolute_error: 40.1072 - mean_absolute_percentage_error: 44.9714\n", "Epoch 33/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2776.7515 - mean_absolute_error: 39.4257 - mean_absolute_percentage_error: 43.9125\n", "Epoch 34/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2729.4502 - mean_absolute_error: 38.8133 - mean_absolute_percentage_error: 42.9989\n", "Epoch 35/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 2687.1089 - mean_absolute_error: 38.2454 - mean_absolute_percentage_error: 42.1472\n", "Epoch 36/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 2649.5278 - mean_absolute_error: 37.7127 - mean_absolute_percentage_error: 41.3561\n", "Epoch 37/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2616.1899 - mean_absolute_error: 37.1526 - mean_absolute_percentage_error: 40.5163\n", "Epoch 38/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2586.3806 - mean_absolute_error: 36.5741 - mean_absolute_percentage_error: 39.6310\n", "Epoch 39/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2559.2861 - mean_absolute_error: 36.0075 - mean_absolute_percentage_error: 38.7477\n", "Epoch 40/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2534.0837 - mean_absolute_error: 35.4278 - mean_absolute_percentage_error: 37.8246\n", "Epoch 41/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 2510.0239 - mean_absolute_error: 34.8468 - mean_absolute_percentage_error: 36.8864\n", "Epoch 42/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 2486.4934 - mean_absolute_error: 34.2624 - mean_absolute_percentage_error: 35.9279\n", "Epoch 43/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 2463.0508 - mean_absolute_error: 33.6739 - mean_absolute_percentage_error: 34.9485\n", "Epoch 44/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 2439.4390 - mean_absolute_error: 33.0895 - mean_absolute_percentage_error: 33.9618\n", "Epoch 45/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 2415.5669 - mean_absolute_error: 32.5090 - mean_absolute_percentage_error: 32.9688\n", "Epoch 46/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2391.4814 - mean_absolute_error: 31.9338 - mean_absolute_percentage_error: 31.9774\n", "Epoch 47/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 2367.3354 - mean_absolute_error: 31.3679 - mean_absolute_percentage_error: 30.9978\n", "Epoch 48/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 2343.3499 - mean_absolute_error: 30.8207 - mean_absolute_percentage_error: 30.0451\n", "Epoch 49/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 2319.7788 - mean_absolute_error: 30.2940 - mean_absolute_percentage_error: 29.1237\n", "Epoch 50/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 2296.8662 - mean_absolute_error: 29.8026 - mean_absolute_percentage_error: 28.2631\n", "Epoch 51/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 2274.8123 - mean_absolute_error: 29.3515 - mean_absolute_percentage_error: 27.4735\n", "Epoch 52/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2253.7522 - mean_absolute_error: 28.9391 - mean_absolute_percentage_error: 26.7555\n", "Epoch 53/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2233.7529 - mean_absolute_error: 28.5568 - mean_absolute_percentage_error: 26.0963\n", "Epoch 54/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2214.8257 - mean_absolute_error: 28.2158 - mean_absolute_percentage_error: 25.5193\n", "Epoch 55/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2196.9402 - mean_absolute_error: 27.9185 - mean_absolute_percentage_error: 25.0297\n", "Epoch 56/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 2180.0408 - mean_absolute_error: 27.6761 - mean_absolute_percentage_error: 24.6479\n", "Epoch 57/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 2164.0576 - mean_absolute_error: 27.4878 - mean_absolute_percentage_error: 24.3808\n", "Epoch 58/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2148.9143 - mean_absolute_error: 27.3373 - mean_absolute_percentage_error: 24.1920\n", "Epoch 59/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2134.5330 - mean_absolute_error: 27.2238 - mean_absolute_percentage_error: 24.0754\n", "Epoch 60/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2120.8347 - mean_absolute_error: 27.1335 - mean_absolute_percentage_error: 24.0081\n", "Epoch 61/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2107.7424 - mean_absolute_error: 27.0742 - mean_absolute_percentage_error: 24.0079\n", "Epoch 62/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2095.1790 - mean_absolute_error: 27.0347 - mean_absolute_percentage_error: 24.0501\n", "Epoch 63/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 2083.0693 - mean_absolute_error: 27.0129 - mean_absolute_percentage_error: 24.1288\n", "Epoch 64/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2071.3408 - mean_absolute_error: 27.0026 - mean_absolute_percentage_error: 24.2305\n", "Epoch 65/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2059.9260 - mean_absolute_error: 26.9991 - mean_absolute_percentage_error: 24.3398\n", "Epoch 66/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2048.7632 - mean_absolute_error: 27.0001 - mean_absolute_percentage_error: 24.4519\n", "Epoch 67/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 2037.8013 - mean_absolute_error: 26.9954 - mean_absolute_percentage_error: 24.5502\n", "Epoch 68/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2026.9994 - mean_absolute_error: 26.9837 - mean_absolute_percentage_error: 24.6311\n", "Epoch 69/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 2016.3281 - mean_absolute_error: 26.9628 - mean_absolute_percentage_error: 24.6929\n", "Epoch 70/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 2005.7695 - mean_absolute_error: 26.9356 - mean_absolute_percentage_error: 24.7408\n", "Epoch 71/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1995.3167 - mean_absolute_error: 26.8999 - mean_absolute_percentage_error: 24.7702\n", "Epoch 72/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1984.9713 - mean_absolute_error: 26.8541 - mean_absolute_percentage_error: 24.7773\n", "Epoch 73/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1974.7428 - mean_absolute_error: 26.8009 - mean_absolute_percentage_error: 24.7665\n", "Epoch 74/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1964.6447 - mean_absolute_error: 26.7404 - mean_absolute_percentage_error: 24.7373\n", "Epoch 75/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1954.6948 - mean_absolute_error: 26.6723 - mean_absolute_percentage_error: 24.6903\n", "Epoch 76/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1944.9108 - mean_absolute_error: 26.5991 - mean_absolute_percentage_error: 24.6294\n", "Epoch 77/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1935.3080 - mean_absolute_error: 26.5224 - mean_absolute_percentage_error: 24.5565\n", "Epoch 78/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1925.8993 - mean_absolute_error: 26.4424 - mean_absolute_percentage_error: 24.4736\n", "Epoch 79/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 1916.6927 - mean_absolute_error: 26.3598 - mean_absolute_percentage_error: 24.3821\n", "Epoch 80/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1907.6896 - mean_absolute_error: 26.2761 - mean_absolute_percentage_error: 24.2844\n", "Epoch 81/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 1898.8862 - mean_absolute_error: 26.1926 - mean_absolute_percentage_error: 24.1828\n", "Epoch 82/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1890.2749 - mean_absolute_error: 26.1095 - mean_absolute_percentage_error: 24.0786\n", "Epoch 83/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1881.8419 - mean_absolute_error: 26.0271 - mean_absolute_percentage_error: 23.9727\n", "Epoch 84/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1873.5731 - mean_absolute_error: 25.9455 - mean_absolute_percentage_error: 23.8663\n", "Epoch 85/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1865.4525 - mean_absolute_error: 25.8658 - mean_absolute_percentage_error: 23.7617\n", "Epoch 86/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1857.4648 - mean_absolute_error: 25.7881 - mean_absolute_percentage_error: 23.6595\n", "Epoch 87/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1849.5972 - mean_absolute_error: 25.7121 - mean_absolute_percentage_error: 23.5598\n", "Epoch 88/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1841.8380 - mean_absolute_error: 25.6381 - mean_absolute_percentage_error: 23.4632\n", "Epoch 89/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1834.1805 - mean_absolute_error: 25.5667 - mean_absolute_percentage_error: 23.3706\n", "Epoch 90/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1826.6193 - mean_absolute_error: 25.4970 - mean_absolute_percentage_error: 23.2819\n", "Epoch 91/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1819.1515 - mean_absolute_error: 25.4301 - mean_absolute_percentage_error: 23.1987\n", "Epoch 92/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1811.7760 - mean_absolute_error: 25.3653 - mean_absolute_percentage_error: 23.1199\n", "Epoch 93/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1804.4922 - mean_absolute_error: 25.3030 - mean_absolute_percentage_error: 23.0463\n", "Epoch 94/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1797.2990 - mean_absolute_error: 25.2429 - mean_absolute_percentage_error: 22.9767\n", "Epoch 95/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1790.1954 - mean_absolute_error: 25.1857 - mean_absolute_percentage_error: 22.9123\n", "Epoch 96/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1783.1794 - mean_absolute_error: 25.1295 - mean_absolute_percentage_error: 22.8506\n", "Epoch 97/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1776.2469 - mean_absolute_error: 25.0745 - mean_absolute_percentage_error: 22.7915\n", "Epoch 98/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1769.3944 - mean_absolute_error: 25.0200 - mean_absolute_percentage_error: 22.7348\n", "Epoch 99/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1762.6155 - mean_absolute_error: 24.9659 - mean_absolute_percentage_error: 22.6800\n", "Epoch 100/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1755.9044 - mean_absolute_error: 24.9122 - mean_absolute_percentage_error: 22.6268\n", "Epoch 101/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1749.2542 - mean_absolute_error: 24.8590 - mean_absolute_percentage_error: 22.5760\n", "Epoch 102/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1742.6576 - mean_absolute_error: 24.8055 - mean_absolute_percentage_error: 22.5260\n", "Epoch 103/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1736.1080 - mean_absolute_error: 24.7517 - mean_absolute_percentage_error: 22.4767\n", "Epoch 104/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1729.5995 - mean_absolute_error: 24.6986 - mean_absolute_percentage_error: 22.4283\n", "Epoch 105/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1723.1268 - mean_absolute_error: 24.6455 - mean_absolute_percentage_error: 22.3807\n", "Epoch 106/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1716.6857 - mean_absolute_error: 24.5923 - mean_absolute_percentage_error: 22.3339\n", "Epoch 107/1000\n", "1/1 [==============================] - 0s 10ms/step - loss: 1710.2736 - mean_absolute_error: 24.5388 - mean_absolute_percentage_error: 22.2874\n", "Epoch 108/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1703.8892 - mean_absolute_error: 24.4850 - mean_absolute_percentage_error: 22.2414\n", "Epoch 109/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1697.5326 - mean_absolute_error: 24.4311 - mean_absolute_percentage_error: 22.1959\n", "Epoch 110/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1691.2056 - mean_absolute_error: 24.3779 - mean_absolute_percentage_error: 22.1524\n", "Epoch 111/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1684.9108 - mean_absolute_error: 24.3254 - mean_absolute_percentage_error: 22.1108\n", "Epoch 112/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1678.6523 - mean_absolute_error: 24.2747 - mean_absolute_percentage_error: 22.0720\n", "Epoch 113/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1672.4344 - mean_absolute_error: 24.2252 - mean_absolute_percentage_error: 22.0353\n", "Epoch 114/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1666.2626 - mean_absolute_error: 24.1762 - mean_absolute_percentage_error: 21.9998\n", "Epoch 115/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1660.1412 - mean_absolute_error: 24.1281 - mean_absolute_percentage_error: 21.9658\n", "Epoch 116/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1654.0743 - mean_absolute_error: 24.0817 - mean_absolute_percentage_error: 21.9346\n", "Epoch 117/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1648.0662 - mean_absolute_error: 24.0372 - mean_absolute_percentage_error: 21.9064\n", "Epoch 118/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1642.1180 - mean_absolute_error: 23.9942 - mean_absolute_percentage_error: 21.8799\n", "Epoch 119/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1636.2311 - mean_absolute_error: 23.9531 - mean_absolute_percentage_error: 21.8554\n", "Epoch 120/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1630.4060 - mean_absolute_error: 23.9147 - mean_absolute_percentage_error: 21.8335\n", "Epoch 121/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1624.6411 - mean_absolute_error: 23.8773 - mean_absolute_percentage_error: 21.8128\n", "Epoch 122/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1618.9353 - mean_absolute_error: 23.8405 - mean_absolute_percentage_error: 21.7931\n", "Epoch 123/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1613.2864 - mean_absolute_error: 23.8054 - mean_absolute_percentage_error: 21.7761\n", "Epoch 124/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1607.6921 - mean_absolute_error: 23.7712 - mean_absolute_percentage_error: 21.7601\n", "Epoch 125/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1602.1510 - mean_absolute_error: 23.7374 - mean_absolute_percentage_error: 21.7447\n", "Epoch 126/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1596.6610 - mean_absolute_error: 23.7040 - mean_absolute_percentage_error: 21.7294\n", "Epoch 127/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1591.2213 - mean_absolute_error: 23.6704 - mean_absolute_percentage_error: 21.7139\n", "Epoch 128/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1585.8313 - mean_absolute_error: 23.6374 - mean_absolute_percentage_error: 21.6987\n", "Epoch 129/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1580.4919 - mean_absolute_error: 23.6042 - mean_absolute_percentage_error: 21.6828\n", "Epoch 130/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1575.2046 - mean_absolute_error: 23.5705 - mean_absolute_percentage_error: 21.6662\n", "Epoch 131/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1569.9725 - mean_absolute_error: 23.5363 - mean_absolute_percentage_error: 21.6487\n", "Epoch 132/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1564.7983 - mean_absolute_error: 23.5018 - mean_absolute_percentage_error: 21.6302\n", "Epoch 133/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1559.6866 - mean_absolute_error: 23.4668 - mean_absolute_percentage_error: 21.6106\n", "Epoch 134/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1554.6414 - mean_absolute_error: 23.4317 - mean_absolute_percentage_error: 21.5903\n", "Epoch 135/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1549.6661 - mean_absolute_error: 23.3967 - mean_absolute_percentage_error: 21.5695\n", "Epoch 136/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1544.7649 - mean_absolute_error: 23.3613 - mean_absolute_percentage_error: 21.5477\n", "Epoch 137/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1539.9396 - mean_absolute_error: 23.3262 - mean_absolute_percentage_error: 21.5256\n", "Epoch 138/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1535.1917 - mean_absolute_error: 23.2914 - mean_absolute_percentage_error: 21.5030\n", "Epoch 139/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1530.5210 - mean_absolute_error: 23.2581 - mean_absolute_percentage_error: 21.4815\n", "Epoch 140/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1525.9272 - mean_absolute_error: 23.2260 - mean_absolute_percentage_error: 21.4609\n", "Epoch 141/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1521.4083 - mean_absolute_error: 23.1942 - mean_absolute_percentage_error: 21.4394\n", "Epoch 142/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1516.9623 - mean_absolute_error: 23.1632 - mean_absolute_percentage_error: 21.4184\n", "Epoch 143/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1512.5865 - mean_absolute_error: 23.1329 - mean_absolute_percentage_error: 21.3976\n", "Epoch 144/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1508.2776 - mean_absolute_error: 23.1026 - mean_absolute_percentage_error: 21.3760\n", "Epoch 145/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1504.0331 - mean_absolute_error: 23.0723 - mean_absolute_percentage_error: 21.3537\n", "Epoch 146/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1499.8491 - mean_absolute_error: 23.0422 - mean_absolute_percentage_error: 21.3310\n", "Epoch 147/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1495.7229 - mean_absolute_error: 23.0120 - mean_absolute_percentage_error: 21.3077\n", "Epoch 148/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1491.6515 - mean_absolute_error: 22.9818 - mean_absolute_percentage_error: 21.2838\n", "Epoch 149/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1487.6310 - mean_absolute_error: 22.9516 - mean_absolute_percentage_error: 21.2594\n", "Epoch 150/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1483.6593 - mean_absolute_error: 22.9216 - mean_absolute_percentage_error: 21.2349\n", "Epoch 151/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1479.7332 - mean_absolute_error: 22.8915 - mean_absolute_percentage_error: 21.2098\n", "Epoch 152/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1475.8502 - mean_absolute_error: 22.8620 - mean_absolute_percentage_error: 21.1851\n", "Epoch 153/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1472.0084 - mean_absolute_error: 22.8337 - mean_absolute_percentage_error: 21.1615\n", "Epoch 154/1000\n", "1/1 [==============================] - 0s 14ms/step - loss: 1468.2058 - mean_absolute_error: 22.8057 - mean_absolute_percentage_error: 21.1377\n", "Epoch 155/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1464.4408 - mean_absolute_error: 22.7774 - mean_absolute_percentage_error: 21.1133\n", "Epoch 156/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1460.7122 - mean_absolute_error: 22.7487 - mean_absolute_percentage_error: 21.0883\n", "Epoch 157/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1457.0189 - mean_absolute_error: 22.7199 - mean_absolute_percentage_error: 21.0629\n", "Epoch 158/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1453.3601 - mean_absolute_error: 22.6908 - mean_absolute_percentage_error: 21.0369\n", "Epoch 159/1000\n", "1/1 [==============================] - 0s 12ms/step - loss: 1449.7352 - mean_absolute_error: 22.6616 - mean_absolute_percentage_error: 21.0108\n", "Epoch 160/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1446.1433 - mean_absolute_error: 22.6324 - mean_absolute_percentage_error: 20.9846\n", "Epoch 161/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1442.5839 - mean_absolute_error: 22.6033 - mean_absolute_percentage_error: 20.9589\n", "Epoch 162/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1439.0565 - mean_absolute_error: 22.5747 - mean_absolute_percentage_error: 20.9338\n", "Epoch 163/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1435.5598 - mean_absolute_error: 22.5459 - mean_absolute_percentage_error: 20.9087\n", "Epoch 164/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1432.0938 - mean_absolute_error: 22.5178 - mean_absolute_percentage_error: 20.8842\n", "Epoch 165/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1428.6569 - mean_absolute_error: 22.4899 - mean_absolute_percentage_error: 20.8598\n", "Epoch 166/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1425.2482 - mean_absolute_error: 22.4620 - mean_absolute_percentage_error: 20.8354\n", "Epoch 167/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1421.8663 - mean_absolute_error: 22.4343 - mean_absolute_percentage_error: 20.8111\n", "Epoch 168/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 1418.5098 - mean_absolute_error: 22.4066 - mean_absolute_percentage_error: 20.7868\n", "Epoch 169/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 1415.1765 - mean_absolute_error: 22.3790 - mean_absolute_percentage_error: 20.7625\n", "Epoch 170/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1411.8647 - mean_absolute_error: 22.3517 - mean_absolute_percentage_error: 20.7391\n", "Epoch 171/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1408.5720 - mean_absolute_error: 22.3245 - mean_absolute_percentage_error: 20.7156\n", "Epoch 172/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1405.2957 - mean_absolute_error: 22.2971 - mean_absolute_percentage_error: 20.6919\n", "Epoch 173/1000\n", "1/1 [==============================] - 0s 10ms/step - loss: 1402.0333 - mean_absolute_error: 22.2698 - mean_absolute_percentage_error: 20.6681\n", "Epoch 174/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1398.7819 - mean_absolute_error: 22.2425 - mean_absolute_percentage_error: 20.6443\n", "Epoch 175/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 1395.5388 - mean_absolute_error: 22.2149 - mean_absolute_percentage_error: 20.6200\n", "Epoch 176/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1392.3013 - mean_absolute_error: 22.1870 - mean_absolute_percentage_error: 20.5950\n", "Epoch 177/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1389.0669 - mean_absolute_error: 22.1592 - mean_absolute_percentage_error: 20.5698\n", "Epoch 178/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1385.8326 - mean_absolute_error: 22.1315 - mean_absolute_percentage_error: 20.5447\n", "Epoch 179/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1382.5964 - mean_absolute_error: 22.1035 - mean_absolute_percentage_error: 20.5191\n", "Epoch 180/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1379.3563 - mean_absolute_error: 22.0750 - mean_absolute_percentage_error: 20.4925\n", "Epoch 181/1000\n", "1/1 [==============================] - 0s 13ms/step - loss: 1376.1106 - mean_absolute_error: 22.0457 - mean_absolute_percentage_error: 20.4647\n", "Epoch 182/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1372.8577 - mean_absolute_error: 22.0165 - mean_absolute_percentage_error: 20.4366\n", "Epoch 183/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1369.5968 - mean_absolute_error: 21.9876 - mean_absolute_percentage_error: 20.4086\n", "Epoch 184/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1366.3274 - mean_absolute_error: 21.9580 - mean_absolute_percentage_error: 20.3793\n", "Epoch 185/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1363.0496 - mean_absolute_error: 21.9278 - mean_absolute_percentage_error: 20.3487\n", "Epoch 186/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 1359.7635 - mean_absolute_error: 21.8971 - mean_absolute_percentage_error: 20.3172\n", "Epoch 187/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1356.4705 - mean_absolute_error: 21.8658 - mean_absolute_percentage_error: 20.2846\n", "Epoch 188/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1353.1716 - mean_absolute_error: 21.8340 - mean_absolute_percentage_error: 20.2509\n", "Epoch 189/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 1349.8690 - mean_absolute_error: 21.8014 - mean_absolute_percentage_error: 20.2158\n", "Epoch 190/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1346.5649 - mean_absolute_error: 21.7680 - mean_absolute_percentage_error: 20.1794\n", "Epoch 191/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1343.2618 - mean_absolute_error: 21.7340 - mean_absolute_percentage_error: 20.1418\n", "Epoch 192/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1339.9626 - mean_absolute_error: 21.6994 - mean_absolute_percentage_error: 20.1033\n", "Epoch 193/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 1336.6704 - mean_absolute_error: 21.6647 - mean_absolute_percentage_error: 20.0642\n", "Epoch 194/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1333.3881 - mean_absolute_error: 21.6298 - mean_absolute_percentage_error: 20.0251\n", "Epoch 195/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 1330.1185 - mean_absolute_error: 21.5957 - mean_absolute_percentage_error: 19.9874\n", "Epoch 196/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1326.8647 - mean_absolute_error: 21.5626 - mean_absolute_percentage_error: 19.9516\n", "Epoch 197/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1323.6292 - mean_absolute_error: 21.5295 - mean_absolute_percentage_error: 19.9162\n", "Epoch 198/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1320.4141 - mean_absolute_error: 21.4966 - mean_absolute_percentage_error: 19.8809\n", "Epoch 199/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1317.2218 - mean_absolute_error: 21.4639 - mean_absolute_percentage_error: 19.8461\n", "Epoch 200/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1314.0538 - mean_absolute_error: 21.4315 - mean_absolute_percentage_error: 19.8119\n", "Epoch 201/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1310.9119 - mean_absolute_error: 21.4000 - mean_absolute_percentage_error: 19.7793\n", "Epoch 202/1000\n", "1/1 [==============================] - 0s 980us/step - loss: 1307.7966 - mean_absolute_error: 21.3692 - mean_absolute_percentage_error: 19.7481\n", "Epoch 203/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1304.7095 - mean_absolute_error: 21.3393 - mean_absolute_percentage_error: 19.7194\n", "Epoch 204/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1301.6505 - mean_absolute_error: 21.3108 - mean_absolute_percentage_error: 19.6929\n", "Epoch 205/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1298.6205 - mean_absolute_error: 21.2829 - mean_absolute_percentage_error: 19.6677\n", "Epoch 206/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1295.6191 - mean_absolute_error: 21.2564 - mean_absolute_percentage_error: 19.6449\n", "Epoch 207/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1292.6466 - mean_absolute_error: 21.2310 - mean_absolute_percentage_error: 19.6242\n", "Epoch 208/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1289.7025 - mean_absolute_error: 21.2065 - mean_absolute_percentage_error: 19.6050\n", "Epoch 209/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1286.7860 - mean_absolute_error: 21.1830 - mean_absolute_percentage_error: 19.5878\n", "Epoch 210/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1283.8970 - mean_absolute_error: 21.1604 - mean_absolute_percentage_error: 19.5725\n", "Epoch 211/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1281.0343 - mean_absolute_error: 21.1387 - mean_absolute_percentage_error: 19.5587\n", "Epoch 212/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1278.1968 - mean_absolute_error: 21.1186 - mean_absolute_percentage_error: 19.5482\n", "Epoch 213/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1275.3834 - mean_absolute_error: 21.0991 - mean_absolute_percentage_error: 19.5389\n", "Epoch 214/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1272.5933 - mean_absolute_error: 21.0801 - mean_absolute_percentage_error: 19.5305\n", "Epoch 215/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1269.8248 - mean_absolute_error: 21.0614 - mean_absolute_percentage_error: 19.5227\n", "Epoch 216/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1267.0769 - mean_absolute_error: 21.0429 - mean_absolute_percentage_error: 19.5155\n", "Epoch 217/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1264.3484 - mean_absolute_error: 21.0250 - mean_absolute_percentage_error: 19.5092\n", "Epoch 218/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1261.6381 - mean_absolute_error: 21.0074 - mean_absolute_percentage_error: 19.5035\n", "Epoch 219/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1258.9451 - mean_absolute_error: 20.9902 - mean_absolute_percentage_error: 19.4983\n", "Epoch 220/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1256.2683 - mean_absolute_error: 20.9732 - mean_absolute_percentage_error: 19.4936\n", "Epoch 221/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1253.6077 - mean_absolute_error: 20.9561 - mean_absolute_percentage_error: 19.4888\n", "Epoch 222/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1250.9626 - mean_absolute_error: 20.9390 - mean_absolute_percentage_error: 19.4840\n", "Epoch 223/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1248.3328 - mean_absolute_error: 20.9219 - mean_absolute_percentage_error: 19.4792\n", "Epoch 224/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1245.7183 - mean_absolute_error: 20.9046 - mean_absolute_percentage_error: 19.4741\n", "Epoch 225/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1243.1194 - mean_absolute_error: 20.8874 - mean_absolute_percentage_error: 19.4690\n", "Epoch 226/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1240.5364 - mean_absolute_error: 20.8701 - mean_absolute_percentage_error: 19.4637\n", "Epoch 227/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1237.9696 - mean_absolute_error: 20.8533 - mean_absolute_percentage_error: 19.4587\n", "Epoch 228/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1235.4199 - mean_absolute_error: 20.8361 - mean_absolute_percentage_error: 19.4533\n", "Epoch 229/1000\n", "1/1 [==============================] - 0s 11ms/step - loss: 1232.8876 - mean_absolute_error: 20.8190 - mean_absolute_percentage_error: 19.4474\n", "Epoch 230/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 1230.3734 - mean_absolute_error: 20.8018 - mean_absolute_percentage_error: 19.4410\n", "Epoch 231/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 1227.8778 - mean_absolute_error: 20.7845 - mean_absolute_percentage_error: 19.4342\n", "Epoch 232/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1225.4012 - mean_absolute_error: 20.7671 - mean_absolute_percentage_error: 19.4271\n", "Epoch 233/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1222.9442 - mean_absolute_error: 20.7495 - mean_absolute_percentage_error: 19.4193\n", "Epoch 234/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1220.5068 - mean_absolute_error: 20.7316 - mean_absolute_percentage_error: 19.4109\n", "Epoch 235/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1218.0894 - mean_absolute_error: 20.7135 - mean_absolute_percentage_error: 19.4019\n", "Epoch 236/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1215.6914 - mean_absolute_error: 20.6956 - mean_absolute_percentage_error: 19.3925\n", "Epoch 237/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1213.3134 - mean_absolute_error: 20.6776 - mean_absolute_percentage_error: 19.3825\n", "Epoch 238/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1210.9546 - mean_absolute_error: 20.6594 - mean_absolute_percentage_error: 19.3720\n", "Epoch 239/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1208.6147 - mean_absolute_error: 20.6416 - mean_absolute_percentage_error: 19.3615\n", "Epoch 240/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1206.2933 - mean_absolute_error: 20.6235 - mean_absolute_percentage_error: 19.3507\n", "Epoch 241/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1203.9900 - mean_absolute_error: 20.6054 - mean_absolute_percentage_error: 19.3393\n", "Epoch 242/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1201.7039 - mean_absolute_error: 20.5870 - mean_absolute_percentage_error: 19.3273\n", "Epoch 243/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 1199.4348 - mean_absolute_error: 20.5688 - mean_absolute_percentage_error: 19.3153\n", "Epoch 244/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 1197.1818 - mean_absolute_error: 20.5506 - mean_absolute_percentage_error: 19.3028\n", "Epoch 245/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1194.9446 - mean_absolute_error: 20.5322 - mean_absolute_percentage_error: 19.2898\n", "Epoch 246/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1192.7224 - mean_absolute_error: 20.5138 - mean_absolute_percentage_error: 19.2764\n", "Epoch 247/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 1190.5151 - mean_absolute_error: 20.4953 - mean_absolute_percentage_error: 19.2626\n", "Epoch 248/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1188.3220 - mean_absolute_error: 20.4767 - mean_absolute_percentage_error: 19.2485\n", "Epoch 249/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 1186.1427 - mean_absolute_error: 20.4585 - mean_absolute_percentage_error: 19.2346\n", "Epoch 250/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1183.9771 - mean_absolute_error: 20.4408 - mean_absolute_percentage_error: 19.2210\n", "Epoch 251/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1181.8248 - mean_absolute_error: 20.4229 - mean_absolute_percentage_error: 19.2070\n", "Epoch 252/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1179.6858 - mean_absolute_error: 20.4049 - mean_absolute_percentage_error: 19.1925\n", "Epoch 253/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1177.5598 - mean_absolute_error: 20.3869 - mean_absolute_percentage_error: 19.1778\n", "Epoch 254/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1175.4468 - mean_absolute_error: 20.3695 - mean_absolute_percentage_error: 19.1633\n", "Epoch 255/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1173.3466 - mean_absolute_error: 20.3524 - mean_absolute_percentage_error: 19.1488\n", "Epoch 256/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1171.2595 - mean_absolute_error: 20.3355 - mean_absolute_percentage_error: 19.1343\n", "Epoch 257/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1169.1848 - mean_absolute_error: 20.3186 - mean_absolute_percentage_error: 19.1197\n", "Epoch 258/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1167.1234 - mean_absolute_error: 20.3020 - mean_absolute_percentage_error: 19.1056\n", "Epoch 259/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1165.0745 - mean_absolute_error: 20.2855 - mean_absolute_percentage_error: 19.0915\n", "Epoch 260/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1163.0388 - mean_absolute_error: 20.2692 - mean_absolute_percentage_error: 19.0776\n", "Epoch 261/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1161.0156 - mean_absolute_error: 20.2528 - mean_absolute_percentage_error: 19.0633\n", "Epoch 262/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1159.0054 - mean_absolute_error: 20.2366 - mean_absolute_percentage_error: 19.0493\n", "Epoch 263/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1157.0082 - mean_absolute_error: 20.2205 - mean_absolute_percentage_error: 19.0355\n", "Epoch 264/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1155.0237 - mean_absolute_error: 20.2046 - mean_absolute_percentage_error: 19.0217\n", "Epoch 265/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1153.0522 - mean_absolute_error: 20.1885 - mean_absolute_percentage_error: 19.0076\n", "Epoch 266/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1151.0939 - mean_absolute_error: 20.1724 - mean_absolute_percentage_error: 18.9933\n", "Epoch 267/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1149.1486 - mean_absolute_error: 20.1563 - mean_absolute_percentage_error: 18.9790\n", "Epoch 268/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1147.2166 - mean_absolute_error: 20.1408 - mean_absolute_percentage_error: 18.9652\n", "Epoch 269/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1145.2980 - mean_absolute_error: 20.1253 - mean_absolute_percentage_error: 18.9512\n", "Epoch 270/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1143.3926 - mean_absolute_error: 20.1096 - mean_absolute_percentage_error: 18.9371\n", "Epoch 271/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1141.5012 - mean_absolute_error: 20.0945 - mean_absolute_percentage_error: 18.9233\n", "Epoch 272/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1139.6235 - mean_absolute_error: 20.0795 - mean_absolute_percentage_error: 18.9097\n", "Epoch 273/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1137.7595 - mean_absolute_error: 20.0645 - mean_absolute_percentage_error: 18.8961\n", "Epoch 274/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1135.9098 - mean_absolute_error: 20.0501 - mean_absolute_percentage_error: 18.8832\n", "Epoch 275/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1134.0737 - mean_absolute_error: 20.0362 - mean_absolute_percentage_error: 18.8712\n", "Epoch 276/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1132.2520 - mean_absolute_error: 20.0226 - mean_absolute_percentage_error: 18.8597\n", "Epoch 277/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1130.4441 - mean_absolute_error: 20.0091 - mean_absolute_percentage_error: 18.8482\n", "Epoch 278/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1128.6499 - mean_absolute_error: 19.9957 - mean_absolute_percentage_error: 18.8370\n", "Epoch 279/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1126.8696 - mean_absolute_error: 19.9825 - mean_absolute_percentage_error: 18.8261\n", "Epoch 280/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1125.1028 - mean_absolute_error: 19.9694 - mean_absolute_percentage_error: 18.8156\n", "Epoch 281/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1123.3491 - mean_absolute_error: 19.9563 - mean_absolute_percentage_error: 18.8050\n", "Epoch 282/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1121.6086 - mean_absolute_error: 19.9432 - mean_absolute_percentage_error: 18.7944\n", "Epoch 283/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1119.8810 - mean_absolute_error: 19.9302 - mean_absolute_percentage_error: 18.7839\n", "Epoch 284/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1118.1653 - mean_absolute_error: 19.9171 - mean_absolute_percentage_error: 18.7734\n", "Epoch 285/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1116.4620 - mean_absolute_error: 19.9042 - mean_absolute_percentage_error: 18.7632\n", "Epoch 286/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1114.7703 - mean_absolute_error: 19.8918 - mean_absolute_percentage_error: 18.7539\n", "Epoch 287/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1113.0898 - mean_absolute_error: 19.8796 - mean_absolute_percentage_error: 18.7451\n", "Epoch 288/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1111.4204 - mean_absolute_error: 19.8675 - mean_absolute_percentage_error: 18.7363\n", "Epoch 289/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1109.7615 - mean_absolute_error: 19.8554 - mean_absolute_percentage_error: 18.7278\n", "Epoch 290/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1108.1128 - mean_absolute_error: 19.8436 - mean_absolute_percentage_error: 18.7198\n", "Epoch 291/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1106.4739 - mean_absolute_error: 19.8326 - mean_absolute_percentage_error: 18.7130\n", "Epoch 292/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1104.8446 - mean_absolute_error: 19.8219 - mean_absolute_percentage_error: 18.7068\n", "Epoch 293/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1103.2245 - mean_absolute_error: 19.8114 - mean_absolute_percentage_error: 18.7008\n", "Epoch 294/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 1101.6130 - mean_absolute_error: 19.8009 - mean_absolute_percentage_error: 18.6948\n", "Epoch 295/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1100.0103 - mean_absolute_error: 19.7903 - mean_absolute_percentage_error: 18.6888\n", "Epoch 296/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1098.4156 - mean_absolute_error: 19.7797 - mean_absolute_percentage_error: 18.6829\n", "Epoch 297/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1096.8292 - mean_absolute_error: 19.7692 - mean_absolute_percentage_error: 18.6771\n", "Epoch 298/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1095.2505 - mean_absolute_error: 19.7587 - mean_absolute_percentage_error: 18.6713\n", "Epoch 299/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1093.6796 - mean_absolute_error: 19.7481 - mean_absolute_percentage_error: 18.6654\n", "Epoch 300/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1092.1165 - mean_absolute_error: 19.7376 - mean_absolute_percentage_error: 18.6598\n", "Epoch 301/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 1090.5612 - mean_absolute_error: 19.7273 - mean_absolute_percentage_error: 18.6544\n", "Epoch 302/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1089.0135 - mean_absolute_error: 19.7170 - mean_absolute_percentage_error: 18.6491\n", "Epoch 303/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1087.4738 - mean_absolute_error: 19.7065 - mean_absolute_percentage_error: 18.6437\n", "Epoch 304/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1085.9420 - mean_absolute_error: 19.6960 - mean_absolute_percentage_error: 18.6382\n", "Epoch 305/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1084.4185 - mean_absolute_error: 19.6854 - mean_absolute_percentage_error: 18.6325\n", "Epoch 306/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1082.9036 - mean_absolute_error: 19.6747 - mean_absolute_percentage_error: 18.6267\n", "Epoch 307/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1081.3969 - mean_absolute_error: 19.6642 - mean_absolute_percentage_error: 18.6214\n", "Epoch 308/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1079.8988 - mean_absolute_error: 19.6536 - mean_absolute_percentage_error: 18.6158\n", "Epoch 309/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1078.4098 - mean_absolute_error: 19.6429 - mean_absolute_percentage_error: 18.6101\n", "Epoch 310/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1076.9294 - mean_absolute_error: 19.6320 - mean_absolute_percentage_error: 18.6041\n", "Epoch 311/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1075.4579 - mean_absolute_error: 19.6210 - mean_absolute_percentage_error: 18.5980\n", "Epoch 312/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1073.9952 - mean_absolute_error: 19.6099 - mean_absolute_percentage_error: 18.5917\n", "Epoch 313/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1072.5415 - mean_absolute_error: 19.5989 - mean_absolute_percentage_error: 18.5854\n", "Epoch 314/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 1071.0961 - mean_absolute_error: 19.5878 - mean_absolute_percentage_error: 18.5790\n", "Epoch 315/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 1069.6594 - mean_absolute_error: 19.5767 - mean_absolute_percentage_error: 18.5725\n", "Epoch 316/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1068.2312 - mean_absolute_error: 19.5658 - mean_absolute_percentage_error: 18.5660\n", "Epoch 317/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1066.8112 - mean_absolute_error: 19.5548 - mean_absolute_percentage_error: 18.5593\n", "Epoch 318/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 1065.3992 - mean_absolute_error: 19.5437 - mean_absolute_percentage_error: 18.5524\n", "Epoch 319/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1063.9949 - mean_absolute_error: 19.5325 - mean_absolute_percentage_error: 18.5455\n", "Epoch 320/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1062.5983 - mean_absolute_error: 19.5212 - mean_absolute_percentage_error: 18.5384\n", "Epoch 321/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1061.2091 - mean_absolute_error: 19.5098 - mean_absolute_percentage_error: 18.5311\n", "Epoch 322/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1059.8270 - mean_absolute_error: 19.4984 - mean_absolute_percentage_error: 18.5238\n", "Epoch 323/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 1058.4519 - mean_absolute_error: 19.4871 - mean_absolute_percentage_error: 18.5167\n", "Epoch 324/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1057.0835 - mean_absolute_error: 19.4759 - mean_absolute_percentage_error: 18.5096\n", "Epoch 325/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 1055.7216 - mean_absolute_error: 19.4646 - mean_absolute_percentage_error: 18.5023\n", "Epoch 326/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1054.3658 - mean_absolute_error: 19.4532 - mean_absolute_percentage_error: 18.4949\n", "Epoch 327/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1053.0160 - mean_absolute_error: 19.4417 - mean_absolute_percentage_error: 18.4874\n", "Epoch 328/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1051.6719 - mean_absolute_error: 19.4301 - mean_absolute_percentage_error: 18.4797\n", "Epoch 329/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1050.3330 - mean_absolute_error: 19.4185 - mean_absolute_percentage_error: 18.4719\n", "Epoch 330/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1048.9991 - mean_absolute_error: 19.4068 - mean_absolute_percentage_error: 18.4640\n", "Epoch 331/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1047.6699 - mean_absolute_error: 19.3950 - mean_absolute_percentage_error: 18.4559\n", "Epoch 332/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1046.3450 - mean_absolute_error: 19.3831 - mean_absolute_percentage_error: 18.4478\n", "Epoch 333/1000\n", "1/1 [==============================] - 0s 13ms/step - loss: 1045.0237 - mean_absolute_error: 19.3713 - mean_absolute_percentage_error: 18.4398\n", "Epoch 334/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1043.7061 - mean_absolute_error: 19.3600 - mean_absolute_percentage_error: 18.4321\n", "Epoch 335/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1042.3917 - mean_absolute_error: 19.3486 - mean_absolute_percentage_error: 18.4244\n", "Epoch 336/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1041.0798 - mean_absolute_error: 19.3376 - mean_absolute_percentage_error: 18.4170\n", "Epoch 337/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1039.7708 - mean_absolute_error: 19.3267 - mean_absolute_percentage_error: 18.4095\n", "Epoch 338/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1038.4640 - mean_absolute_error: 19.3158 - mean_absolute_percentage_error: 18.4021\n", "Epoch 339/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1037.1594 - mean_absolute_error: 19.3049 - mean_absolute_percentage_error: 18.3946\n", "Epoch 340/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1035.8574 - mean_absolute_error: 19.2939 - mean_absolute_percentage_error: 18.3871\n", "Epoch 341/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1034.5579 - mean_absolute_error: 19.2831 - mean_absolute_percentage_error: 18.3798\n", "Epoch 342/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1033.2617 - mean_absolute_error: 19.2724 - mean_absolute_percentage_error: 18.3727\n", "Epoch 343/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1031.9694 - mean_absolute_error: 19.2619 - mean_absolute_percentage_error: 18.3657\n", "Epoch 344/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1030.6819 - mean_absolute_error: 19.2517 - mean_absolute_percentage_error: 18.3591\n", "Epoch 345/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1029.4004 - mean_absolute_error: 19.2417 - mean_absolute_percentage_error: 18.3526\n", "Epoch 346/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1028.1262 - mean_absolute_error: 19.2320 - mean_absolute_percentage_error: 18.3465\n", "Epoch 347/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1026.8606 - mean_absolute_error: 19.2225 - mean_absolute_percentage_error: 18.3406\n", "Epoch 348/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1025.6051 - mean_absolute_error: 19.2133 - mean_absolute_percentage_error: 18.3352\n", "Epoch 349/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1024.3608 - mean_absolute_error: 19.2043 - mean_absolute_percentage_error: 18.3298\n", "Epoch 350/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1023.1289 - mean_absolute_error: 19.1955 - mean_absolute_percentage_error: 18.3246\n", "Epoch 351/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1021.9102 - mean_absolute_error: 19.1868 - mean_absolute_percentage_error: 18.3196\n", "Epoch 352/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1020.7051 - mean_absolute_error: 19.1784 - mean_absolute_percentage_error: 18.3148\n", "Epoch 353/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1019.5138 - mean_absolute_error: 19.1701 - mean_absolute_percentage_error: 18.3100\n", "Epoch 354/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1018.3365 - mean_absolute_error: 19.1619 - mean_absolute_percentage_error: 18.3053\n", "Epoch 355/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1017.1726 - mean_absolute_error: 19.1538 - mean_absolute_percentage_error: 18.3006\n", "Epoch 356/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 1016.0215 - mean_absolute_error: 19.1460 - mean_absolute_percentage_error: 18.2961\n", "Epoch 357/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1014.8828 - mean_absolute_error: 19.1383 - mean_absolute_percentage_error: 18.2915\n", "Epoch 358/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1013.7556 - mean_absolute_error: 19.1306 - mean_absolute_percentage_error: 18.2870\n", "Epoch 359/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1012.6390 - mean_absolute_error: 19.1230 - mean_absolute_percentage_error: 18.2823\n", "Epoch 360/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1011.5325 - mean_absolute_error: 19.1153 - mean_absolute_percentage_error: 18.2773\n", "Epoch 361/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1010.4352 - mean_absolute_error: 19.1075 - mean_absolute_percentage_error: 18.2722\n", "Epoch 362/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1009.3464 - mean_absolute_error: 19.0996 - mean_absolute_percentage_error: 18.2668\n", "Epoch 363/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1008.2659 - mean_absolute_error: 19.0915 - mean_absolute_percentage_error: 18.2611\n", "Epoch 364/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1007.1927 - mean_absolute_error: 19.0834 - mean_absolute_percentage_error: 18.2551\n", "Epoch 365/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1006.1263 - mean_absolute_error: 19.0752 - mean_absolute_percentage_error: 18.2489\n", "Epoch 366/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1005.0663 - mean_absolute_error: 19.0668 - mean_absolute_percentage_error: 18.2423\n", "Epoch 367/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 1004.0120 - mean_absolute_error: 19.0584 - mean_absolute_percentage_error: 18.2357\n", "Epoch 368/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1002.9628 - mean_absolute_error: 19.0500 - mean_absolute_percentage_error: 18.2288\n", "Epoch 369/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 1001.9181 - mean_absolute_error: 19.0414 - mean_absolute_percentage_error: 18.2217\n", "Epoch 370/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 1000.8773 - mean_absolute_error: 19.0326 - mean_absolute_percentage_error: 18.2142\n", "Epoch 371/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 999.8400 - mean_absolute_error: 19.0237 - mean_absolute_percentage_error: 18.2064\n", "Epoch 372/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 998.8053 - mean_absolute_error: 19.0145 - mean_absolute_percentage_error: 18.1983\n", "Epoch 373/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 997.7725 - mean_absolute_error: 19.0053 - mean_absolute_percentage_error: 18.1899\n", "Epoch 374/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 996.7415 - mean_absolute_error: 18.9958 - mean_absolute_percentage_error: 18.1812\n", "Epoch 375/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 995.7110 - mean_absolute_error: 18.9862 - mean_absolute_percentage_error: 18.1723\n", "Epoch 376/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 994.6810 - mean_absolute_error: 18.9765 - mean_absolute_percentage_error: 18.1632\n", "Epoch 377/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 993.6507 - mean_absolute_error: 18.9667 - mean_absolute_percentage_error: 18.1538\n", "Epoch 378/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 992.6201 - mean_absolute_error: 18.9567 - mean_absolute_percentage_error: 18.1441\n", "Epoch 379/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 991.5884 - mean_absolute_error: 18.9464 - mean_absolute_percentage_error: 18.1341\n", "Epoch 380/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 990.5554 - mean_absolute_error: 18.9360 - mean_absolute_percentage_error: 18.1237\n", "Epoch 381/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 989.5211 - mean_absolute_error: 18.9254 - mean_absolute_percentage_error: 18.1130\n", "Epoch 382/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 988.4851 - mean_absolute_error: 18.9145 - mean_absolute_percentage_error: 18.1019\n", "Epoch 383/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 987.4478 - mean_absolute_error: 18.9036 - mean_absolute_percentage_error: 18.0907\n", "Epoch 384/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 986.4092 - mean_absolute_error: 18.8924 - mean_absolute_percentage_error: 18.0791\n", "Epoch 385/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 985.3696 - mean_absolute_error: 18.8811 - mean_absolute_percentage_error: 18.0671\n", "Epoch 386/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 984.3293 - mean_absolute_error: 18.8695 - mean_absolute_percentage_error: 18.0547\n", "Epoch 387/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 983.2892 - mean_absolute_error: 18.8577 - mean_absolute_percentage_error: 18.0420\n", "Epoch 388/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 982.2500 - mean_absolute_error: 18.8463 - mean_absolute_percentage_error: 18.0298\n", "Epoch 389/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 981.2125 - mean_absolute_error: 18.8348 - mean_absolute_percentage_error: 18.0176\n", "Epoch 390/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 980.1774 - mean_absolute_error: 18.8232 - mean_absolute_percentage_error: 18.0051\n", "Epoch 391/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 979.1465 - mean_absolute_error: 18.8115 - mean_absolute_percentage_error: 17.9924\n", "Epoch 392/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 978.1207 - mean_absolute_error: 18.8002 - mean_absolute_percentage_error: 17.9801\n", "Epoch 393/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 977.1015 - mean_absolute_error: 18.7891 - mean_absolute_percentage_error: 17.9680\n", "Epoch 394/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 976.0900 - mean_absolute_error: 18.7781 - mean_absolute_percentage_error: 17.9558\n", "Epoch 395/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 975.0880 - mean_absolute_error: 18.7673 - mean_absolute_percentage_error: 17.9440\n", "Epoch 396/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 974.0968 - mean_absolute_error: 18.7568 - mean_absolute_percentage_error: 17.9325\n", "Epoch 397/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 973.1173 - mean_absolute_error: 18.7467 - mean_absolute_percentage_error: 17.9218\n", "Epoch 398/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 972.1511 - mean_absolute_error: 18.7369 - mean_absolute_percentage_error: 17.9116\n", "Epoch 399/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 971.1986 - mean_absolute_error: 18.7273 - mean_absolute_percentage_error: 17.9018\n", "Epoch 400/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 970.2605 - mean_absolute_error: 18.7180 - mean_absolute_percentage_error: 17.8926\n", "Epoch 401/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 969.3375 - mean_absolute_error: 18.7090 - mean_absolute_percentage_error: 17.8840\n", "Epoch 402/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 968.4293 - mean_absolute_error: 18.7003 - mean_absolute_percentage_error: 17.8761\n", "Epoch 403/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 967.5356 - mean_absolute_error: 18.6920 - mean_absolute_percentage_error: 17.8689\n", "Epoch 404/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 966.6562 - mean_absolute_error: 18.6841 - mean_absolute_percentage_error: 17.8624\n", "Epoch 405/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 965.7901 - mean_absolute_error: 18.6766 - mean_absolute_percentage_error: 17.8566\n", "Epoch 406/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 964.9366 - mean_absolute_error: 18.6696 - mean_absolute_percentage_error: 17.8517\n", "Epoch 407/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 964.0947 - mean_absolute_error: 18.6629 - mean_absolute_percentage_error: 17.8475\n", "Epoch 408/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 963.2631 - mean_absolute_error: 18.6565 - mean_absolute_percentage_error: 17.8439\n", "Epoch 409/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 962.4409 - mean_absolute_error: 18.6506 - mean_absolute_percentage_error: 17.8410\n", "Epoch 410/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 961.6270 - mean_absolute_error: 18.6451 - mean_absolute_percentage_error: 17.8388\n", "Epoch 411/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 960.8203 - mean_absolute_error: 18.6399 - mean_absolute_percentage_error: 17.8371\n", "Epoch 412/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 960.0201 - mean_absolute_error: 18.6348 - mean_absolute_percentage_error: 17.8357\n", "Epoch 413/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 959.2256 - mean_absolute_error: 18.6302 - mean_absolute_percentage_error: 17.8349\n", "Epoch 414/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 958.4362 - mean_absolute_error: 18.6258 - mean_absolute_percentage_error: 17.8344\n", "Epoch 415/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 957.6511 - mean_absolute_error: 18.6216 - mean_absolute_percentage_error: 17.8341\n", "Epoch 416/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 956.8702 - mean_absolute_error: 18.6174 - mean_absolute_percentage_error: 17.8340\n", "Epoch 417/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 956.0928 - mean_absolute_error: 18.6132 - mean_absolute_percentage_error: 17.8339\n", "Epoch 418/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 955.3190 - mean_absolute_error: 18.6091 - mean_absolute_percentage_error: 17.8339\n", "Epoch 419/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 954.5485 - mean_absolute_error: 18.6050 - mean_absolute_percentage_error: 17.8340\n", "Epoch 420/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 953.7809 - mean_absolute_error: 18.6009 - mean_absolute_percentage_error: 17.8340\n", "Epoch 421/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 953.0166 - mean_absolute_error: 18.5966 - mean_absolute_percentage_error: 17.8338\n", "Epoch 422/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 952.2552 - mean_absolute_error: 18.5923 - mean_absolute_percentage_error: 17.8335\n", "Epoch 423/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 951.4969 - mean_absolute_error: 18.5878 - mean_absolute_percentage_error: 17.8329\n", "Epoch 424/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 950.7416 - mean_absolute_error: 18.5832 - mean_absolute_percentage_error: 17.8320\n", "Epoch 425/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 949.9892 - mean_absolute_error: 18.5785 - mean_absolute_percentage_error: 17.8309\n", "Epoch 426/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 949.2398 - mean_absolute_error: 18.5736 - mean_absolute_percentage_error: 17.8296\n", "Epoch 427/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 948.4934 - mean_absolute_error: 18.5686 - mean_absolute_percentage_error: 17.8279\n", "Epoch 428/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 947.7500 - mean_absolute_error: 18.5635 - mean_absolute_percentage_error: 17.8260\n", "Epoch 429/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 947.0095 - mean_absolute_error: 18.5583 - mean_absolute_percentage_error: 17.8240\n", "Epoch 430/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 946.2721 - mean_absolute_error: 18.5531 - mean_absolute_percentage_error: 17.8218\n", "Epoch 431/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 945.5375 - mean_absolute_error: 18.5478 - mean_absolute_percentage_error: 17.8195\n", "Epoch 432/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 944.8060 - mean_absolute_error: 18.5425 - mean_absolute_percentage_error: 17.8172\n", "Epoch 433/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 944.0775 - mean_absolute_error: 18.5371 - mean_absolute_percentage_error: 17.8146\n", "Epoch 434/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 943.3519 - mean_absolute_error: 18.5316 - mean_absolute_percentage_error: 17.8119\n", "Epoch 435/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 942.6292 - mean_absolute_error: 18.5260 - mean_absolute_percentage_error: 17.8089\n", "Epoch 436/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 941.9094 - mean_absolute_error: 18.5204 - mean_absolute_percentage_error: 17.8058\n", "Epoch 437/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 941.1926 - mean_absolute_error: 18.5147 - mean_absolute_percentage_error: 17.8026\n", "Epoch 438/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 940.4785 - mean_absolute_error: 18.5090 - mean_absolute_percentage_error: 17.7992\n", "Epoch 439/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 939.7673 - mean_absolute_error: 18.5033 - mean_absolute_percentage_error: 17.7958\n", "Epoch 440/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 939.0590 - mean_absolute_error: 18.4975 - mean_absolute_percentage_error: 17.7922\n", "Epoch 441/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 938.3535 - mean_absolute_error: 18.4918 - mean_absolute_percentage_error: 17.7885\n", "Epoch 442/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 937.6505 - mean_absolute_error: 18.4860 - mean_absolute_percentage_error: 17.7848\n", "Epoch 443/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 936.9503 - mean_absolute_error: 18.4801 - mean_absolute_percentage_error: 17.7809\n", "Epoch 444/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 936.2529 - mean_absolute_error: 18.4743 - mean_absolute_percentage_error: 17.7770\n", "Epoch 445/1000\n", "1/1 [==============================] - 0s 11ms/step - loss: 935.5582 - mean_absolute_error: 18.4684 - mean_absolute_percentage_error: 17.7730\n", "Epoch 446/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 934.8658 - mean_absolute_error: 18.4625 - mean_absolute_percentage_error: 17.7689\n", "Epoch 447/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 934.1762 - mean_absolute_error: 18.4566 - mean_absolute_percentage_error: 17.7648\n", "Epoch 448/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 933.4891 - mean_absolute_error: 18.4508 - mean_absolute_percentage_error: 17.7607\n", "Epoch 449/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 932.8046 - mean_absolute_error: 18.4449 - mean_absolute_percentage_error: 17.7566\n", "Epoch 450/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 932.1223 - mean_absolute_error: 18.4390 - mean_absolute_percentage_error: 17.7525\n", "Epoch 451/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 931.4425 - mean_absolute_error: 18.4332 - mean_absolute_percentage_error: 17.7484\n", "Epoch 452/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 930.7653 - mean_absolute_error: 18.4275 - mean_absolute_percentage_error: 17.7444\n", "Epoch 453/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 930.0903 - mean_absolute_error: 18.4219 - mean_absolute_percentage_error: 17.7405\n", "Epoch 454/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 929.4177 - mean_absolute_error: 18.4163 - mean_absolute_percentage_error: 17.7367\n", "Epoch 455/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 928.7473 - mean_absolute_error: 18.4108 - mean_absolute_percentage_error: 17.7329\n", "Epoch 456/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 928.0793 - mean_absolute_error: 18.4055 - mean_absolute_percentage_error: 17.7294\n", "Epoch 457/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 927.4135 - mean_absolute_error: 18.4002 - mean_absolute_percentage_error: 17.7259\n", "Epoch 458/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 926.7500 - mean_absolute_error: 18.3949 - mean_absolute_percentage_error: 17.7224\n", "Epoch 459/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 926.0887 - mean_absolute_error: 18.3896 - mean_absolute_percentage_error: 17.7189\n", "Epoch 460/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 925.4296 - mean_absolute_error: 18.3843 - mean_absolute_percentage_error: 17.7154\n", "Epoch 461/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 924.7725 - mean_absolute_error: 18.3790 - mean_absolute_percentage_error: 17.7119\n", "Epoch 462/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 924.1179 - mean_absolute_error: 18.3738 - mean_absolute_percentage_error: 17.7084\n", "Epoch 463/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 923.4653 - mean_absolute_error: 18.3685 - mean_absolute_percentage_error: 17.7049\n", "Epoch 464/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 922.8148 - mean_absolute_error: 18.3632 - mean_absolute_percentage_error: 17.7013\n", "Epoch 465/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 922.1663 - mean_absolute_error: 18.3578 - mean_absolute_percentage_error: 17.6978\n", "Epoch 466/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 921.5203 - mean_absolute_error: 18.3525 - mean_absolute_percentage_error: 17.6942\n", "Epoch 467/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 920.8760 - mean_absolute_error: 18.3472 - mean_absolute_percentage_error: 17.6906\n", "Epoch 468/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 920.2341 - mean_absolute_error: 18.3419 - mean_absolute_percentage_error: 17.6871\n", "Epoch 469/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 919.5942 - mean_absolute_error: 18.3367 - mean_absolute_percentage_error: 17.6836\n", "Epoch 470/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 918.9564 - mean_absolute_error: 18.3314 - mean_absolute_percentage_error: 17.6801\n", "Epoch 471/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 918.3207 - mean_absolute_error: 18.3262 - mean_absolute_percentage_error: 17.6765\n", "Epoch 472/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 917.6869 - mean_absolute_error: 18.3209 - mean_absolute_percentage_error: 17.6729\n", "Epoch 473/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 917.0553 - mean_absolute_error: 18.3156 - mean_absolute_percentage_error: 17.6693\n", "Epoch 474/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 916.4258 - mean_absolute_error: 18.3104 - mean_absolute_percentage_error: 17.6658\n", "Epoch 475/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 915.7984 - mean_absolute_error: 18.3055 - mean_absolute_percentage_error: 17.6624\n", "Epoch 476/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 915.1729 - mean_absolute_error: 18.3006 - mean_absolute_percentage_error: 17.6590\n", "Epoch 477/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 914.5496 - mean_absolute_error: 18.2958 - mean_absolute_percentage_error: 17.6558\n", "Epoch 478/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 913.9283 - mean_absolute_error: 18.2910 - mean_absolute_percentage_error: 17.6526\n", "Epoch 479/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 913.3090 - mean_absolute_error: 18.2863 - mean_absolute_percentage_error: 17.6494\n", "Epoch 480/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 912.6918 - mean_absolute_error: 18.2815 - mean_absolute_percentage_error: 17.6461\n", "Epoch 481/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 912.0765 - mean_absolute_error: 18.2767 - mean_absolute_percentage_error: 17.6428\n", "Epoch 482/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 911.4634 - mean_absolute_error: 18.2721 - mean_absolute_percentage_error: 17.6397\n", "Epoch 483/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 910.8523 - mean_absolute_error: 18.2676 - mean_absolute_percentage_error: 17.6366\n", "Epoch 484/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 910.2431 - mean_absolute_error: 18.2631 - mean_absolute_percentage_error: 17.6335\n", "Epoch 485/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 909.6359 - mean_absolute_error: 18.2586 - mean_absolute_percentage_error: 17.6305\n", "Epoch 486/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 909.0308 - mean_absolute_error: 18.2541 - mean_absolute_percentage_error: 17.6274\n", "Epoch 487/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 908.4275 - mean_absolute_error: 18.2496 - mean_absolute_percentage_error: 17.6242\n", "Epoch 488/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 907.8263 - mean_absolute_error: 18.2451 - mean_absolute_percentage_error: 17.6211\n", "Epoch 489/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 907.2269 - mean_absolute_error: 18.2405 - mean_absolute_percentage_error: 17.6179\n", "Epoch 490/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 906.6295 - mean_absolute_error: 18.2360 - mean_absolute_percentage_error: 17.6147\n", "Epoch 491/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 906.0341 - mean_absolute_error: 18.2314 - mean_absolute_percentage_error: 17.6114\n", "Epoch 492/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 905.4404 - mean_absolute_error: 18.2269 - mean_absolute_percentage_error: 17.6082\n", "Epoch 493/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 904.8487 - mean_absolute_error: 18.2223 - mean_absolute_percentage_error: 17.6051\n", "Epoch 494/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 904.2589 - mean_absolute_error: 18.2178 - mean_absolute_percentage_error: 17.6019\n", "Epoch 495/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 903.6708 - mean_absolute_error: 18.2132 - mean_absolute_percentage_error: 17.5987\n", "Epoch 496/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 903.0844 - mean_absolute_error: 18.2087 - mean_absolute_percentage_error: 17.5955\n", "Epoch 497/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 902.5000 - mean_absolute_error: 18.2041 - mean_absolute_percentage_error: 17.5922\n", "Epoch 498/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 901.9171 - mean_absolute_error: 18.1995 - mean_absolute_percentage_error: 17.5890\n", "Epoch 499/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 901.3361 - mean_absolute_error: 18.1949 - mean_absolute_percentage_error: 17.5857\n", "Epoch 500/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 900.7568 - mean_absolute_error: 18.1903 - mean_absolute_percentage_error: 17.5826\n", "Epoch 501/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 900.1790 - mean_absolute_error: 18.1858 - mean_absolute_percentage_error: 17.5794\n", "Epoch 502/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 899.6028 - mean_absolute_error: 18.1812 - mean_absolute_percentage_error: 17.5762\n", "Epoch 503/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 899.0283 - mean_absolute_error: 18.1768 - mean_absolute_percentage_error: 17.5733\n", "Epoch 504/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 898.4553 - mean_absolute_error: 18.1724 - mean_absolute_percentage_error: 17.5704\n", "Epoch 505/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 897.8839 - mean_absolute_error: 18.1679 - mean_absolute_percentage_error: 17.5674\n", "Epoch 506/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 897.3137 - mean_absolute_error: 18.1635 - mean_absolute_percentage_error: 17.5645\n", "Epoch 507/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 896.7451 - mean_absolute_error: 18.1591 - mean_absolute_percentage_error: 17.5616\n", "Epoch 508/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 896.1778 - mean_absolute_error: 18.1547 - mean_absolute_percentage_error: 17.5586\n", "Epoch 509/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 895.6119 - mean_absolute_error: 18.1502 - mean_absolute_percentage_error: 17.5556\n", "Epoch 510/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 895.0472 - mean_absolute_error: 18.1457 - mean_absolute_percentage_error: 17.5526\n", "Epoch 511/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 894.4837 - mean_absolute_error: 18.1413 - mean_absolute_percentage_error: 17.5496\n", "Epoch 512/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 893.9213 - mean_absolute_error: 18.1367 - mean_absolute_percentage_error: 17.5465\n", "Epoch 513/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 893.3602 - mean_absolute_error: 18.1322 - mean_absolute_percentage_error: 17.5434\n", "Epoch 514/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 892.8000 - mean_absolute_error: 18.1276 - mean_absolute_percentage_error: 17.5401\n", "Epoch 515/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 892.2409 - mean_absolute_error: 18.1229 - mean_absolute_percentage_error: 17.5368\n", "Epoch 516/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 891.6829 - mean_absolute_error: 18.1183 - mean_absolute_percentage_error: 17.5334\n", "Epoch 517/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 891.1254 - mean_absolute_error: 18.1136 - mean_absolute_percentage_error: 17.5299\n", "Epoch 518/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 890.5690 - mean_absolute_error: 18.1091 - mean_absolute_percentage_error: 17.5266\n", "Epoch 519/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 890.0134 - mean_absolute_error: 18.1045 - mean_absolute_percentage_error: 17.5231\n", "Epoch 520/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 889.4580 - mean_absolute_error: 18.0999 - mean_absolute_percentage_error: 17.5196\n", "Epoch 521/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 888.9037 - mean_absolute_error: 18.0952 - mean_absolute_percentage_error: 17.5160\n", "Epoch 522/1000\n", "1/1 [==============================] - 0s 10ms/step - loss: 888.3497 - mean_absolute_error: 18.0904 - mean_absolute_percentage_error: 17.5123\n", "Epoch 523/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 887.7961 - mean_absolute_error: 18.0857 - mean_absolute_percentage_error: 17.5085\n", "Epoch 524/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 887.2431 - mean_absolute_error: 18.0808 - mean_absolute_percentage_error: 17.5046\n", "Epoch 525/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 886.6902 - mean_absolute_error: 18.0759 - mean_absolute_percentage_error: 17.5006\n", "Epoch 526/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 886.1375 - mean_absolute_error: 18.0710 - mean_absolute_percentage_error: 17.4965\n", "Epoch 527/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 885.5850 - mean_absolute_error: 18.0660 - mean_absolute_percentage_error: 17.4923\n", "Epoch 528/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 885.0325 - mean_absolute_error: 18.0610 - mean_absolute_percentage_error: 17.4881\n", "Epoch 529/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 884.4800 - mean_absolute_error: 18.0558 - mean_absolute_percentage_error: 17.4837\n", "Epoch 530/1000\n", "1/1 [==============================] - 0s 10ms/step - loss: 883.9273 - mean_absolute_error: 18.0507 - mean_absolute_percentage_error: 17.4792\n", "Epoch 531/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 883.3744 - mean_absolute_error: 18.0454 - mean_absolute_percentage_error: 17.4746\n", "Epoch 532/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 882.8212 - mean_absolute_error: 18.0401 - mean_absolute_percentage_error: 17.4699\n", "Epoch 533/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 882.2676 - mean_absolute_error: 18.0348 - mean_absolute_percentage_error: 17.4651\n", "Epoch 534/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 881.7136 - mean_absolute_error: 18.0293 - mean_absolute_percentage_error: 17.4601\n", "Epoch 535/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 881.1591 - mean_absolute_error: 18.0238 - mean_absolute_percentage_error: 17.4550\n", "Epoch 536/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 880.6038 - mean_absolute_error: 18.0182 - mean_absolute_percentage_error: 17.4499\n", "Epoch 537/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 880.0478 - mean_absolute_error: 18.0126 - mean_absolute_percentage_error: 17.4445\n", "Epoch 538/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 879.4912 - mean_absolute_error: 18.0068 - mean_absolute_percentage_error: 17.4391\n", "Epoch 539/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 878.9337 - mean_absolute_error: 18.0010 - mean_absolute_percentage_error: 17.4336\n", "Epoch 540/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 878.3753 - mean_absolute_error: 17.9953 - mean_absolute_percentage_error: 17.4281\n", "Epoch 541/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 877.8160 - mean_absolute_error: 17.9894 - mean_absolute_percentage_error: 17.4225\n", "Epoch 542/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 877.2559 - mean_absolute_error: 17.9835 - mean_absolute_percentage_error: 17.4169\n", "Epoch 543/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 876.6946 - mean_absolute_error: 17.9775 - mean_absolute_percentage_error: 17.4112\n", "Epoch 544/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 876.1324 - mean_absolute_error: 17.9715 - mean_absolute_percentage_error: 17.4054\n", "Epoch 545/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 875.5691 - mean_absolute_error: 17.9654 - mean_absolute_percentage_error: 17.3995\n", "Epoch 546/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 875.0049 - mean_absolute_error: 17.9592 - mean_absolute_percentage_error: 17.3935\n", "Epoch 547/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 874.4399 - mean_absolute_error: 17.9530 - mean_absolute_percentage_error: 17.3874\n", "Epoch 548/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 873.8738 - mean_absolute_error: 17.9467 - mean_absolute_percentage_error: 17.3812\n", "Epoch 549/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 873.3069 - mean_absolute_error: 17.9403 - mean_absolute_percentage_error: 17.3750\n", "Epoch 550/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 872.7393 - mean_absolute_error: 17.9339 - mean_absolute_percentage_error: 17.3686\n", "Epoch 551/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 872.1711 - mean_absolute_error: 17.9274 - mean_absolute_percentage_error: 17.3622\n", "Epoch 552/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 871.6022 - mean_absolute_error: 17.9208 - mean_absolute_percentage_error: 17.3556\n", "Epoch 553/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 871.0329 - mean_absolute_error: 17.9142 - mean_absolute_percentage_error: 17.3490\n", "Epoch 554/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 870.4632 - mean_absolute_error: 17.9076 - mean_absolute_percentage_error: 17.3423\n", "Epoch 555/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 869.8935 - mean_absolute_error: 17.9008 - mean_absolute_percentage_error: 17.3356\n", "Epoch 556/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 869.3236 - mean_absolute_error: 17.8941 - mean_absolute_percentage_error: 17.3287\n", "Epoch 557/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 868.7539 - mean_absolute_error: 17.8873 - mean_absolute_percentage_error: 17.3220\n", "Epoch 558/1000\n", "1/1 [==============================] - 0s 10ms/step - loss: 868.1845 - mean_absolute_error: 17.8806 - mean_absolute_percentage_error: 17.3153\n", "Epoch 559/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 867.6157 - mean_absolute_error: 17.8740 - mean_absolute_percentage_error: 17.3089\n", "Epoch 560/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 867.0475 - mean_absolute_error: 17.8674 - mean_absolute_percentage_error: 17.3024\n", "Epoch 561/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 866.4800 - mean_absolute_error: 17.8607 - mean_absolute_percentage_error: 17.2958\n", "Epoch 562/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 865.9136 - mean_absolute_error: 17.8543 - mean_absolute_percentage_error: 17.2896\n", "Epoch 563/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 865.3482 - mean_absolute_error: 17.8478 - mean_absolute_percentage_error: 17.2833\n", "Epoch 564/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 864.7841 - mean_absolute_error: 17.8413 - mean_absolute_percentage_error: 17.2770\n", "Epoch 565/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 864.2213 - mean_absolute_error: 17.8349 - mean_absolute_percentage_error: 17.2707\n", "Epoch 566/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 863.6602 - mean_absolute_error: 17.8283 - mean_absolute_percentage_error: 17.2643\n", "Epoch 567/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 863.1006 - mean_absolute_error: 17.8218 - mean_absolute_percentage_error: 17.2579\n", "Epoch 568/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 862.5428 - mean_absolute_error: 17.8152 - mean_absolute_percentage_error: 17.2514\n", "Epoch 569/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 861.9866 - mean_absolute_error: 17.8087 - mean_absolute_percentage_error: 17.2450\n", "Epoch 570/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 861.4325 - mean_absolute_error: 17.8021 - mean_absolute_percentage_error: 17.2386\n", "Epoch 571/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 860.8802 - mean_absolute_error: 17.7956 - mean_absolute_percentage_error: 17.2321\n", "Epoch 572/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 860.3298 - mean_absolute_error: 17.7890 - mean_absolute_percentage_error: 17.2255\n", "Epoch 573/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 859.7815 - mean_absolute_error: 17.7824 - mean_absolute_percentage_error: 17.2190\n", "Epoch 574/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 859.2351 - mean_absolute_error: 17.7759 - mean_absolute_percentage_error: 17.2128\n", "Epoch 575/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 858.6908 - mean_absolute_error: 17.7695 - mean_absolute_percentage_error: 17.2065\n", "Epoch 576/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 858.1486 - mean_absolute_error: 17.7631 - mean_absolute_percentage_error: 17.2003\n", "Epoch 577/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 857.6082 - mean_absolute_error: 17.7566 - mean_absolute_percentage_error: 17.1940\n", "Epoch 578/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 857.0700 - mean_absolute_error: 17.7502 - mean_absolute_percentage_error: 17.1878\n", "Epoch 579/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 856.5338 - mean_absolute_error: 17.7439 - mean_absolute_percentage_error: 17.1817\n", "Epoch 580/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 855.9995 - mean_absolute_error: 17.7377 - mean_absolute_percentage_error: 17.1758\n", "Epoch 581/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 855.4672 - mean_absolute_error: 17.7315 - mean_absolute_percentage_error: 17.1699\n", "Epoch 582/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 854.9369 - mean_absolute_error: 17.7252 - mean_absolute_percentage_error: 17.1640\n", "Epoch 583/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 854.4086 - mean_absolute_error: 17.7190 - mean_absolute_percentage_error: 17.1581\n", "Epoch 584/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 853.8822 - mean_absolute_error: 17.7128 - mean_absolute_percentage_error: 17.1521\n", "Epoch 585/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 853.3577 - mean_absolute_error: 17.7065 - mean_absolute_percentage_error: 17.1462\n", "Epoch 586/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 852.8352 - mean_absolute_error: 17.7003 - mean_absolute_percentage_error: 17.1403\n", "Epoch 587/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 852.3145 - mean_absolute_error: 17.6941 - mean_absolute_percentage_error: 17.1343\n", "Epoch 588/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 851.7957 - mean_absolute_error: 17.6879 - mean_absolute_percentage_error: 17.1283\n", "Epoch 589/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 851.2787 - mean_absolute_error: 17.6817 - mean_absolute_percentage_error: 17.1224\n", "Epoch 590/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 850.7637 - mean_absolute_error: 17.6755 - mean_absolute_percentage_error: 17.1164\n", "Epoch 591/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 850.2507 - mean_absolute_error: 17.6694 - mean_absolute_percentage_error: 17.1104\n", "Epoch 592/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 849.7396 - mean_absolute_error: 17.6633 - mean_absolute_percentage_error: 17.1045\n", "Epoch 593/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 849.2302 - mean_absolute_error: 17.6572 - mean_absolute_percentage_error: 17.0986\n", "Epoch 594/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 848.7228 - mean_absolute_error: 17.6511 - mean_absolute_percentage_error: 17.0927\n", "Epoch 595/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 848.2175 - mean_absolute_error: 17.6451 - mean_absolute_percentage_error: 17.0867\n", "Epoch 596/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 847.7140 - mean_absolute_error: 17.6390 - mean_absolute_percentage_error: 17.0808\n", "Epoch 597/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 847.2125 - mean_absolute_error: 17.6330 - mean_absolute_percentage_error: 17.0750\n", "Epoch 598/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 846.7130 - mean_absolute_error: 17.6271 - mean_absolute_percentage_error: 17.0693\n", "Epoch 599/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 846.2156 - mean_absolute_error: 17.6212 - mean_absolute_percentage_error: 17.0636\n", "Epoch 600/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 845.7203 - mean_absolute_error: 17.6153 - mean_absolute_percentage_error: 17.0580\n", "Epoch 601/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 845.2269 - mean_absolute_error: 17.6095 - mean_absolute_percentage_error: 17.0523\n", "Epoch 602/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 844.7356 - mean_absolute_error: 17.6036 - mean_absolute_percentage_error: 17.0467\n", "Epoch 603/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 844.2465 - mean_absolute_error: 17.5978 - mean_absolute_percentage_error: 17.0411\n", "Epoch 604/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 843.7595 - mean_absolute_error: 17.5920 - mean_absolute_percentage_error: 17.0355\n", "Epoch 605/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 843.2745 - mean_absolute_error: 17.5862 - mean_absolute_percentage_error: 17.0301\n", "Epoch 606/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 842.7917 - mean_absolute_error: 17.5805 - mean_absolute_percentage_error: 17.0246\n", "Epoch 607/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 842.3112 - mean_absolute_error: 17.5748 - mean_absolute_percentage_error: 17.0193\n", "Epoch 608/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 841.8325 - mean_absolute_error: 17.5692 - mean_absolute_percentage_error: 17.0139\n", "Epoch 609/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 841.3560 - mean_absolute_error: 17.5635 - mean_absolute_percentage_error: 17.0086\n", "Epoch 610/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 840.8816 - mean_absolute_error: 17.5579 - mean_absolute_percentage_error: 17.0033\n", "Epoch 611/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 840.4092 - mean_absolute_error: 17.5522 - mean_absolute_percentage_error: 16.9980\n", "Epoch 612/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 839.9389 - mean_absolute_error: 17.5467 - mean_absolute_percentage_error: 16.9928\n", "Epoch 613/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 839.4706 - mean_absolute_error: 17.5411 - mean_absolute_percentage_error: 16.9876\n", "Epoch 614/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 839.0040 - mean_absolute_error: 17.5357 - mean_absolute_percentage_error: 16.9826\n", "Epoch 615/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 838.5397 - mean_absolute_error: 17.5302 - mean_absolute_percentage_error: 16.9775\n", "Epoch 616/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 838.0770 - mean_absolute_error: 17.5248 - mean_absolute_percentage_error: 16.9725\n", "Epoch 617/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 837.6162 - mean_absolute_error: 17.5194 - mean_absolute_percentage_error: 16.9675\n", "Epoch 618/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 837.1572 - mean_absolute_error: 17.5140 - mean_absolute_percentage_error: 16.9626\n", "Epoch 619/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 836.7000 - mean_absolute_error: 17.5088 - mean_absolute_percentage_error: 16.9579\n", "Epoch 620/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 836.2444 - mean_absolute_error: 17.5037 - mean_absolute_percentage_error: 16.9532\n", "Epoch 621/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 835.7906 - mean_absolute_error: 17.4986 - mean_absolute_percentage_error: 16.9487\n", "Epoch 622/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 835.3383 - mean_absolute_error: 17.4937 - mean_absolute_percentage_error: 16.9442\n", "Epoch 623/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 834.8875 - mean_absolute_error: 17.4887 - mean_absolute_percentage_error: 16.9397\n", "Epoch 624/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 834.4382 - mean_absolute_error: 17.4838 - mean_absolute_percentage_error: 16.9353\n", "Epoch 625/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 833.9906 - mean_absolute_error: 17.4789 - mean_absolute_percentage_error: 16.9309\n", "Epoch 626/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 833.5442 - mean_absolute_error: 17.4741 - mean_absolute_percentage_error: 16.9265\n", "Epoch 627/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 833.0992 - mean_absolute_error: 17.4694 - mean_absolute_percentage_error: 16.9222\n", "Epoch 628/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 832.6558 - mean_absolute_error: 17.4647 - mean_absolute_percentage_error: 16.9180\n", "Epoch 629/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 832.2135 - mean_absolute_error: 17.4600 - mean_absolute_percentage_error: 16.9138\n", "Epoch 630/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 831.7725 - mean_absolute_error: 17.4554 - mean_absolute_percentage_error: 16.9096\n", "Epoch 631/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 831.3328 - mean_absolute_error: 17.4508 - mean_absolute_percentage_error: 16.9055\n", "Epoch 632/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 830.8942 - mean_absolute_error: 17.4463 - mean_absolute_percentage_error: 16.9014\n", "Epoch 633/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 830.4570 - mean_absolute_error: 17.4417 - mean_absolute_percentage_error: 16.8973\n", "Epoch 634/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 830.0208 - mean_absolute_error: 17.4371 - mean_absolute_percentage_error: 16.8932\n", "Epoch 635/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 829.5858 - mean_absolute_error: 17.4326 - mean_absolute_percentage_error: 16.8891\n", "Epoch 636/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 829.1516 - mean_absolute_error: 17.4280 - mean_absolute_percentage_error: 16.8850\n", "Epoch 637/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 828.7188 - mean_absolute_error: 17.4235 - mean_absolute_percentage_error: 16.8809\n", "Epoch 638/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 828.2869 - mean_absolute_error: 17.4189 - mean_absolute_percentage_error: 16.8768\n", "Epoch 639/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 827.8558 - mean_absolute_error: 17.4144 - mean_absolute_percentage_error: 16.8727\n", "Epoch 640/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 827.4259 - mean_absolute_error: 17.4098 - mean_absolute_percentage_error: 16.8686\n", "Epoch 641/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 826.9968 - mean_absolute_error: 17.4053 - mean_absolute_percentage_error: 16.8645\n", "Epoch 642/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 826.5688 - mean_absolute_error: 17.4009 - mean_absolute_percentage_error: 16.8604\n", "Epoch 643/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 826.1415 - mean_absolute_error: 17.3964 - mean_absolute_percentage_error: 16.8563\n", "Epoch 644/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 825.7151 - mean_absolute_error: 17.3919 - mean_absolute_percentage_error: 16.8521\n", "Epoch 645/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 825.2897 - mean_absolute_error: 17.3875 - mean_absolute_percentage_error: 16.8481\n", "Epoch 646/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 824.8649 - mean_absolute_error: 17.3832 - mean_absolute_percentage_error: 16.8443\n", "Epoch 647/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 824.4410 - mean_absolute_error: 17.3788 - mean_absolute_percentage_error: 16.8405\n", "Epoch 648/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 824.0177 - mean_absolute_error: 17.3746 - mean_absolute_percentage_error: 16.8368\n", "Epoch 649/1000\n", "1/1 [==============================] - 0s 8ms/step - loss: 823.5952 - mean_absolute_error: 17.3704 - mean_absolute_percentage_error: 16.8331\n", "Epoch 650/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 823.1734 - mean_absolute_error: 17.3661 - mean_absolute_percentage_error: 16.8294\n", "Epoch 651/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 822.7521 - mean_absolute_error: 17.3619 - mean_absolute_percentage_error: 16.8256\n", "Epoch 652/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 822.3317 - mean_absolute_error: 17.3576 - mean_absolute_percentage_error: 16.8219\n", "Epoch 653/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 821.9117 - mean_absolute_error: 17.3533 - mean_absolute_percentage_error: 16.8181\n", "Epoch 654/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 821.4923 - mean_absolute_error: 17.3490 - mean_absolute_percentage_error: 16.8143\n", "Epoch 655/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 821.0735 - mean_absolute_error: 17.3447 - mean_absolute_percentage_error: 16.8105\n", "Epoch 656/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 820.6552 - mean_absolute_error: 17.3404 - mean_absolute_percentage_error: 16.8067\n", "Epoch 657/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 820.2374 - mean_absolute_error: 17.3360 - mean_absolute_percentage_error: 16.8028\n", "Epoch 658/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 819.8200 - mean_absolute_error: 17.3317 - mean_absolute_percentage_error: 16.7989\n", "Epoch 659/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 819.4030 - mean_absolute_error: 17.3273 - mean_absolute_percentage_error: 16.7951\n", "Epoch 660/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 818.9863 - mean_absolute_error: 17.3231 - mean_absolute_percentage_error: 16.7914\n", "Epoch 661/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 818.5700 - mean_absolute_error: 17.3188 - mean_absolute_percentage_error: 16.7876\n", "Epoch 662/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 818.1540 - mean_absolute_error: 17.3145 - mean_absolute_percentage_error: 16.7838\n", "Epoch 663/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 817.7383 - mean_absolute_error: 17.3102 - mean_absolute_percentage_error: 16.7799\n", "Epoch 664/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 817.3229 - mean_absolute_error: 17.3058 - mean_absolute_percentage_error: 16.7761\n", "Epoch 665/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 816.9075 - mean_absolute_error: 17.3015 - mean_absolute_percentage_error: 16.7722\n", "Epoch 666/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 816.4924 - mean_absolute_error: 17.2971 - mean_absolute_percentage_error: 16.7683\n", "Epoch 667/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 816.0774 - mean_absolute_error: 17.2927 - mean_absolute_percentage_error: 16.7643\n", "Epoch 668/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 815.6625 - mean_absolute_error: 17.2883 - mean_absolute_percentage_error: 16.7604\n", "Epoch 669/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 815.2477 - mean_absolute_error: 17.2839 - mean_absolute_percentage_error: 16.7564\n", "Epoch 670/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 814.8328 - mean_absolute_error: 17.2794 - mean_absolute_percentage_error: 16.7524\n", "Epoch 671/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 814.4181 - mean_absolute_error: 17.2749 - mean_absolute_percentage_error: 16.7484\n", "Epoch 672/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 814.0032 - mean_absolute_error: 17.2704 - mean_absolute_percentage_error: 16.7443\n", "Epoch 673/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 813.5883 - mean_absolute_error: 17.2659 - mean_absolute_percentage_error: 16.7402\n", "Epoch 674/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 813.1734 - mean_absolute_error: 17.2614 - mean_absolute_percentage_error: 16.7361\n", "Epoch 675/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 812.7584 - mean_absolute_error: 17.2568 - mean_absolute_percentage_error: 16.7319\n", "Epoch 676/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 812.3432 - mean_absolute_error: 17.2522 - mean_absolute_percentage_error: 16.7278\n", "Epoch 677/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 811.9278 - mean_absolute_error: 17.2477 - mean_absolute_percentage_error: 16.7237\n", "Epoch 678/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 811.5123 - mean_absolute_error: 17.2433 - mean_absolute_percentage_error: 16.7197\n", "Epoch 679/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 811.0967 - mean_absolute_error: 17.2389 - mean_absolute_percentage_error: 16.7157\n", "Epoch 680/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 810.6809 - mean_absolute_error: 17.2344 - mean_absolute_percentage_error: 16.7116\n", "Epoch 681/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 810.2649 - mean_absolute_error: 17.2299 - mean_absolute_percentage_error: 16.7075\n", "Epoch 682/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 809.8487 - mean_absolute_error: 17.2253 - mean_absolute_percentage_error: 16.7034\n", "Epoch 683/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 809.4324 - mean_absolute_error: 17.2208 - mean_absolute_percentage_error: 16.6992\n", "Epoch 684/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 809.0159 - mean_absolute_error: 17.2162 - mean_absolute_percentage_error: 16.6950\n", "Epoch 685/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 808.5993 - mean_absolute_error: 17.2116 - mean_absolute_percentage_error: 16.6908\n", "Epoch 686/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 808.1826 - mean_absolute_error: 17.2070 - mean_absolute_percentage_error: 16.6866\n", "Epoch 687/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 807.7656 - mean_absolute_error: 17.2024 - mean_absolute_percentage_error: 16.6823\n", "Epoch 688/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 807.3486 - mean_absolute_error: 17.1978 - mean_absolute_percentage_error: 16.6781\n", "Epoch 689/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 806.9316 - mean_absolute_error: 17.1932 - mean_absolute_percentage_error: 16.6739\n", "Epoch 690/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 806.5144 - mean_absolute_error: 17.1885 - mean_absolute_percentage_error: 16.6697\n", "Epoch 691/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 806.0975 - mean_absolute_error: 17.1839 - mean_absolute_percentage_error: 16.6655\n", "Epoch 692/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 805.6803 - mean_absolute_error: 17.1792 - mean_absolute_percentage_error: 16.6612\n", "Epoch 693/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 805.2634 - mean_absolute_error: 17.1746 - mean_absolute_percentage_error: 16.6570\n", "Epoch 694/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 804.8463 - mean_absolute_error: 17.1700 - mean_absolute_percentage_error: 16.6528\n", "Epoch 695/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 804.4294 - mean_absolute_error: 17.1654 - mean_absolute_percentage_error: 16.6486\n", "Epoch 696/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 804.0126 - mean_absolute_error: 17.1608 - mean_absolute_percentage_error: 16.6445\n", "Epoch 697/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 803.5958 - mean_absolute_error: 17.1562 - mean_absolute_percentage_error: 16.6403\n", "Epoch 698/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 803.1793 - mean_absolute_error: 17.1517 - mean_absolute_percentage_error: 16.6362\n", "Epoch 699/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 802.7628 - mean_absolute_error: 17.1471 - mean_absolute_percentage_error: 16.6321\n", "Epoch 700/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 802.3464 - mean_absolute_error: 17.1425 - mean_absolute_percentage_error: 16.6280\n", "Epoch 701/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 801.9300 - mean_absolute_error: 17.1379 - mean_absolute_percentage_error: 16.6238\n", "Epoch 702/1000\n", "1/1 [==============================] - 0s 1ms/step - loss: 801.5137 - mean_absolute_error: 17.1333 - mean_absolute_percentage_error: 16.6196\n", "Epoch 703/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 801.0974 - mean_absolute_error: 17.1286 - mean_absolute_percentage_error: 16.6154\n", "Epoch 704/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 800.6807 - mean_absolute_error: 17.1240 - mean_absolute_percentage_error: 16.6113\n", "Epoch 705/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 800.2640 - mean_absolute_error: 17.1193 - mean_absolute_percentage_error: 16.6071\n", "Epoch 706/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 799.8470 - mean_absolute_error: 17.1147 - mean_absolute_percentage_error: 16.6028\n", "Epoch 707/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 799.4294 - mean_absolute_error: 17.1100 - mean_absolute_percentage_error: 16.5986\n", "Epoch 708/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 799.0112 - mean_absolute_error: 17.1052 - mean_absolute_percentage_error: 16.5943\n", "Epoch 709/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 798.5921 - mean_absolute_error: 17.1005 - mean_absolute_percentage_error: 16.5900\n", "Epoch 710/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 798.1720 - mean_absolute_error: 17.0957 - mean_absolute_percentage_error: 16.5856\n", "Epoch 711/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 797.7506 - mean_absolute_error: 17.0908 - mean_absolute_percentage_error: 16.5812\n", "Epoch 712/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 797.3275 - mean_absolute_error: 17.0860 - mean_absolute_percentage_error: 16.5769\n", "Epoch 713/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 796.9026 - mean_absolute_error: 17.0811 - mean_absolute_percentage_error: 16.5725\n", "Epoch 714/1000\n", "1/1 [==============================] - 0s 11ms/step - loss: 796.4753 - mean_absolute_error: 17.0762 - mean_absolute_percentage_error: 16.5681\n", "Epoch 715/1000\n", "1/1 [==============================] - 0s 16ms/step - loss: 796.0452 - mean_absolute_error: 17.0713 - mean_absolute_percentage_error: 16.5637\n", "Epoch 716/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 795.6121 - mean_absolute_error: 17.0663 - mean_absolute_percentage_error: 16.5592\n", "Epoch 717/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 795.1754 - mean_absolute_error: 17.0613 - mean_absolute_percentage_error: 16.5548\n", "Epoch 718/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 794.7344 - mean_absolute_error: 17.0563 - mean_absolute_percentage_error: 16.5504\n", "Epoch 719/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 794.2886 - mean_absolute_error: 17.0512 - mean_absolute_percentage_error: 16.5459\n", "Epoch 720/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 793.8378 - mean_absolute_error: 17.0461 - mean_absolute_percentage_error: 16.5415\n", "Epoch 721/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 793.3810 - mean_absolute_error: 17.0410 - mean_absolute_percentage_error: 16.5370\n", "Epoch 722/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 792.9180 - mean_absolute_error: 17.0358 - mean_absolute_percentage_error: 16.5325\n", "Epoch 723/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 792.4482 - mean_absolute_error: 17.0305 - mean_absolute_percentage_error: 16.5280\n", "Epoch 724/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 791.9711 - mean_absolute_error: 17.0251 - mean_absolute_percentage_error: 16.5235\n", "Epoch 725/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 791.4869 - mean_absolute_error: 17.0198 - mean_absolute_percentage_error: 16.5190\n", "Epoch 726/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 790.9952 - mean_absolute_error: 17.0144 - mean_absolute_percentage_error: 16.5144\n", "Epoch 727/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 790.4963 - mean_absolute_error: 17.0090 - mean_absolute_percentage_error: 16.5099\n", "Epoch 728/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 789.9908 - mean_absolute_error: 17.0036 - mean_absolute_percentage_error: 16.5055\n", "Epoch 729/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 789.4794 - mean_absolute_error: 16.9982 - mean_absolute_percentage_error: 16.5011\n", "Epoch 730/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 788.9633 - mean_absolute_error: 16.9928 - mean_absolute_percentage_error: 16.4967\n", "Epoch 731/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 788.4440 - mean_absolute_error: 16.9875 - mean_absolute_percentage_error: 16.4925\n", "Epoch 732/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 787.9231 - mean_absolute_error: 16.9824 - mean_absolute_percentage_error: 16.4886\n", "Epoch 733/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 787.4024 - mean_absolute_error: 16.9778 - mean_absolute_percentage_error: 16.4851\n", "Epoch 734/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 786.8840 - mean_absolute_error: 16.9734 - mean_absolute_percentage_error: 16.4818\n", "Epoch 735/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 786.3694 - mean_absolute_error: 16.9691 - mean_absolute_percentage_error: 16.4788\n", "Epoch 736/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 785.8604 - mean_absolute_error: 16.9651 - mean_absolute_percentage_error: 16.4760\n", "Epoch 737/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 785.3583 - mean_absolute_error: 16.9613 - mean_absolute_percentage_error: 16.4734\n", "Epoch 738/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 784.8637 - mean_absolute_error: 16.9577 - mean_absolute_percentage_error: 16.4711\n", "Epoch 739/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 784.3773 - mean_absolute_error: 16.9547 - mean_absolute_percentage_error: 16.4691\n", "Epoch 740/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 783.8991 - mean_absolute_error: 16.9518 - mean_absolute_percentage_error: 16.4674\n", "Epoch 741/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 783.4290 - mean_absolute_error: 16.9492 - mean_absolute_percentage_error: 16.4658\n", "Epoch 742/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 782.9665 - mean_absolute_error: 16.9466 - mean_absolute_percentage_error: 16.4644\n", "Epoch 743/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 782.5116 - mean_absolute_error: 16.9442 - mean_absolute_percentage_error: 16.4631\n", "Epoch 744/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 782.0630 - mean_absolute_error: 16.9418 - mean_absolute_percentage_error: 16.4618\n", "Epoch 745/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 781.6207 - mean_absolute_error: 16.9395 - mean_absolute_percentage_error: 16.4606\n", "Epoch 746/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 781.1839 - mean_absolute_error: 16.9371 - mean_absolute_percentage_error: 16.4593\n", "Epoch 747/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 780.7520 - mean_absolute_error: 16.9347 - mean_absolute_percentage_error: 16.4579\n", "Epoch 748/1000\n", "1/1 [==============================] - 0s 11ms/step - loss: 780.3249 - mean_absolute_error: 16.9322 - mean_absolute_percentage_error: 16.4563\n", "Epoch 749/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 779.9020 - mean_absolute_error: 16.9296 - mean_absolute_percentage_error: 16.4547\n", "Epoch 750/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 779.4832 - mean_absolute_error: 16.9270 - mean_absolute_percentage_error: 16.4530\n", "Epoch 751/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 779.0681 - mean_absolute_error: 16.9243 - mean_absolute_percentage_error: 16.4512\n", "Epoch 752/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 778.6567 - mean_absolute_error: 16.9215 - mean_absolute_percentage_error: 16.4492\n", "Epoch 753/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 778.2485 - mean_absolute_error: 16.9186 - mean_absolute_percentage_error: 16.4470\n", "Epoch 754/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 777.8439 - mean_absolute_error: 16.9156 - mean_absolute_percentage_error: 16.4447\n", "Epoch 755/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 777.4423 - mean_absolute_error: 16.9124 - mean_absolute_percentage_error: 16.4422\n", "Epoch 756/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 777.0437 - mean_absolute_error: 16.9091 - mean_absolute_percentage_error: 16.4395\n", "Epoch 757/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 776.6482 - mean_absolute_error: 16.9058 - mean_absolute_percentage_error: 16.4368\n", "Epoch 758/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 776.2555 - mean_absolute_error: 16.9023 - mean_absolute_percentage_error: 16.4339\n", "Epoch 759/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 775.8655 - mean_absolute_error: 16.8987 - mean_absolute_percentage_error: 16.4309\n", "Epoch 760/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 775.4781 - mean_absolute_error: 16.8951 - mean_absolute_percentage_error: 16.4278\n", "Epoch 761/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 775.0932 - mean_absolute_error: 16.8914 - mean_absolute_percentage_error: 16.4246\n", "Epoch 762/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 774.7108 - mean_absolute_error: 16.8876 - mean_absolute_percentage_error: 16.4213\n", "Epoch 763/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 774.3306 - mean_absolute_error: 16.8837 - mean_absolute_percentage_error: 16.4180\n", "Epoch 764/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 773.9526 - mean_absolute_error: 16.8799 - mean_absolute_percentage_error: 16.4147\n", "Epoch 765/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 773.5766 - mean_absolute_error: 16.8759 - mean_absolute_percentage_error: 16.4113\n", "Epoch 766/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 773.2025 - mean_absolute_error: 16.8720 - mean_absolute_percentage_error: 16.4078\n", "Epoch 767/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 772.8303 - mean_absolute_error: 16.8680 - mean_absolute_percentage_error: 16.4044\n", "Epoch 768/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 772.4597 - mean_absolute_error: 16.8639 - mean_absolute_percentage_error: 16.4009\n", "Epoch 769/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 772.0908 - mean_absolute_error: 16.8599 - mean_absolute_percentage_error: 16.3974\n", "Epoch 770/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 771.7233 - mean_absolute_error: 16.8559 - mean_absolute_percentage_error: 16.3939\n", "Epoch 771/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 771.3571 - mean_absolute_error: 16.8519 - mean_absolute_percentage_error: 16.3904\n", "Epoch 772/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 770.9923 - mean_absolute_error: 16.8480 - mean_absolute_percentage_error: 16.3870\n", "Epoch 773/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 770.6287 - mean_absolute_error: 16.8443 - mean_absolute_percentage_error: 16.3838\n", "Epoch 774/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 770.2661 - mean_absolute_error: 16.8405 - mean_absolute_percentage_error: 16.3805\n", "Epoch 775/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 769.9046 - mean_absolute_error: 16.8368 - mean_absolute_percentage_error: 16.3773\n", "Epoch 776/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 769.5440 - mean_absolute_error: 16.8330 - mean_absolute_percentage_error: 16.3741\n", "Epoch 777/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 769.1843 - mean_absolute_error: 16.8293 - mean_absolute_percentage_error: 16.3708\n", "Epoch 778/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 768.8256 - mean_absolute_error: 16.8255 - mean_absolute_percentage_error: 16.3675\n", "Epoch 779/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 768.4675 - mean_absolute_error: 16.8216 - mean_absolute_percentage_error: 16.3642\n", "Epoch 780/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 768.1103 - mean_absolute_error: 16.8178 - mean_absolute_percentage_error: 16.3609\n", "Epoch 781/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 767.7538 - mean_absolute_error: 16.8141 - mean_absolute_percentage_error: 16.3577\n", "Epoch 782/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 767.3980 - mean_absolute_error: 16.8103 - mean_absolute_percentage_error: 16.3544\n", "Epoch 783/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 767.0427 - mean_absolute_error: 16.8065 - mean_absolute_percentage_error: 16.3511\n", "Epoch 784/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 766.6882 - mean_absolute_error: 16.8027 - mean_absolute_percentage_error: 16.3477\n", "Epoch 785/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 766.3343 - mean_absolute_error: 16.7989 - mean_absolute_percentage_error: 16.3444\n", "Epoch 786/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 765.9809 - mean_absolute_error: 16.7950 - mean_absolute_percentage_error: 16.3410\n", "Epoch 787/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 765.6281 - mean_absolute_error: 16.7911 - mean_absolute_percentage_error: 16.3376\n", "Epoch 788/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 765.2758 - mean_absolute_error: 16.7872 - mean_absolute_percentage_error: 16.3341\n", "Epoch 789/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 764.9241 - mean_absolute_error: 16.7833 - mean_absolute_percentage_error: 16.3307\n", "Epoch 790/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 764.5728 - mean_absolute_error: 16.7794 - mean_absolute_percentage_error: 16.3273\n", "Epoch 791/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 764.2221 - mean_absolute_error: 16.7757 - mean_absolute_percentage_error: 16.3242\n", "Epoch 792/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 763.8718 - mean_absolute_error: 16.7719 - mean_absolute_percentage_error: 16.3211\n", "Epoch 793/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 763.5219 - mean_absolute_error: 16.7682 - mean_absolute_percentage_error: 16.3179\n", "Epoch 794/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 763.1726 - mean_absolute_error: 16.7643 - mean_absolute_percentage_error: 16.3147\n", "Epoch 795/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 762.8235 - mean_absolute_error: 16.7605 - mean_absolute_percentage_error: 16.3115\n", "Epoch 796/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 762.4749 - mean_absolute_error: 16.7566 - mean_absolute_percentage_error: 16.3083\n", "Epoch 797/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 762.1268 - mean_absolute_error: 16.7528 - mean_absolute_percentage_error: 16.3051\n", "Epoch 798/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 761.7789 - mean_absolute_error: 16.7490 - mean_absolute_percentage_error: 16.3019\n", "Epoch 799/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 761.4314 - mean_absolute_error: 16.7451 - mean_absolute_percentage_error: 16.2987\n", "Epoch 800/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 761.0844 - mean_absolute_error: 16.7412 - mean_absolute_percentage_error: 16.2955\n", "Epoch 801/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 760.7375 - mean_absolute_error: 16.7373 - mean_absolute_percentage_error: 16.2922\n", "Epoch 802/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 760.3911 - mean_absolute_error: 16.7333 - mean_absolute_percentage_error: 16.2890\n", "Epoch 803/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 760.0449 - mean_absolute_error: 16.7294 - mean_absolute_percentage_error: 16.2857\n", "Epoch 804/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 759.6990 - mean_absolute_error: 16.7254 - mean_absolute_percentage_error: 16.2823\n", "Epoch 805/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 759.3533 - mean_absolute_error: 16.7214 - mean_absolute_percentage_error: 16.2790\n", "Epoch 806/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 759.0079 - mean_absolute_error: 16.7174 - mean_absolute_percentage_error: 16.2756\n", "Epoch 807/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 758.6627 - mean_absolute_error: 16.7134 - mean_absolute_percentage_error: 16.2723\n", "Epoch 808/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 758.3177 - mean_absolute_error: 16.7093 - mean_absolute_percentage_error: 16.2689\n", "Epoch 809/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 757.9730 - mean_absolute_error: 16.7053 - mean_absolute_percentage_error: 16.2654\n", "Epoch 810/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 757.6283 - mean_absolute_error: 16.7012 - mean_absolute_percentage_error: 16.2620\n", "Epoch 811/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 757.2836 - mean_absolute_error: 16.6971 - mean_absolute_percentage_error: 16.2586\n", "Epoch 812/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 756.9391 - mean_absolute_error: 16.6929 - mean_absolute_percentage_error: 16.2551\n", "Epoch 813/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 756.5948 - mean_absolute_error: 16.6888 - mean_absolute_percentage_error: 16.2517\n", "Epoch 814/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 756.2504 - mean_absolute_error: 16.6847 - mean_absolute_percentage_error: 16.2483\n", "Epoch 815/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 755.9059 - mean_absolute_error: 16.6805 - mean_absolute_percentage_error: 16.2448\n", "Epoch 816/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 755.5615 - mean_absolute_error: 16.6764 - mean_absolute_percentage_error: 16.2414\n", "Epoch 817/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 755.2169 - mean_absolute_error: 16.6722 - mean_absolute_percentage_error: 16.2380\n", "Epoch 818/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 754.8722 - mean_absolute_error: 16.6680 - mean_absolute_percentage_error: 16.2345\n", "Epoch 819/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 754.5273 - mean_absolute_error: 16.6638 - mean_absolute_percentage_error: 16.2311\n", "Epoch 820/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 754.1824 - mean_absolute_error: 16.6596 - mean_absolute_percentage_error: 16.2277\n", "Epoch 821/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 753.8369 - mean_absolute_error: 16.6555 - mean_absolute_percentage_error: 16.2244\n", "Epoch 822/1000\n", "1/1 [==============================] - 0s 10ms/step - loss: 753.4915 - mean_absolute_error: 16.6514 - mean_absolute_percentage_error: 16.2211\n", "Epoch 823/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 753.1454 - mean_absolute_error: 16.6473 - mean_absolute_percentage_error: 16.2178\n", "Epoch 824/1000\n", "1/1 [==============================] - 0s 15ms/step - loss: 752.7991 - mean_absolute_error: 16.6431 - mean_absolute_percentage_error: 16.2145\n", "Epoch 825/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 752.4524 - mean_absolute_error: 16.6390 - mean_absolute_percentage_error: 16.2112\n", "Epoch 826/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 752.1050 - mean_absolute_error: 16.6348 - mean_absolute_percentage_error: 16.2079\n", "Epoch 827/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 751.7574 - mean_absolute_error: 16.6307 - mean_absolute_percentage_error: 16.2045\n", "Epoch 828/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 751.4091 - mean_absolute_error: 16.6265 - mean_absolute_percentage_error: 16.2013\n", "Epoch 829/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 751.0601 - mean_absolute_error: 16.6224 - mean_absolute_percentage_error: 16.1980\n", "Epoch 830/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 750.7106 - mean_absolute_error: 16.6182 - mean_absolute_percentage_error: 16.1948\n", "Epoch 831/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 750.3604 - mean_absolute_error: 16.6140 - mean_absolute_percentage_error: 16.1916\n", "Epoch 832/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 750.0095 - mean_absolute_error: 16.6099 - mean_absolute_percentage_error: 16.1886\n", "Epoch 833/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 749.6578 - mean_absolute_error: 16.6058 - mean_absolute_percentage_error: 16.1856\n", "Epoch 834/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 749.3054 - mean_absolute_error: 16.6018 - mean_absolute_percentage_error: 16.1826\n", "Epoch 835/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 748.9522 - mean_absolute_error: 16.5978 - mean_absolute_percentage_error: 16.1797\n", "Epoch 836/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 748.5983 - mean_absolute_error: 16.5938 - mean_absolute_percentage_error: 16.1768\n", "Epoch 837/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 748.2435 - mean_absolute_error: 16.5898 - mean_absolute_percentage_error: 16.1739\n", "Epoch 838/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 747.8881 - mean_absolute_error: 16.5857 - mean_absolute_percentage_error: 16.1710\n", "Epoch 839/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 747.5317 - mean_absolute_error: 16.5816 - mean_absolute_percentage_error: 16.1681\n", "Epoch 840/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 747.1747 - mean_absolute_error: 16.5775 - mean_absolute_percentage_error: 16.1651\n", "Epoch 841/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 746.8170 - mean_absolute_error: 16.5734 - mean_absolute_percentage_error: 16.1622\n", "Epoch 842/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 746.4586 - mean_absolute_error: 16.5693 - mean_absolute_percentage_error: 16.1593\n", "Epoch 843/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 746.0997 - mean_absolute_error: 16.5653 - mean_absolute_percentage_error: 16.1565\n", "Epoch 844/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 745.7402 - mean_absolute_error: 16.5612 - mean_absolute_percentage_error: 16.1537\n", "Epoch 845/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 745.3802 - mean_absolute_error: 16.5573 - mean_absolute_percentage_error: 16.1511\n", "Epoch 846/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 745.0201 - mean_absolute_error: 16.5533 - mean_absolute_percentage_error: 16.1485\n", "Epoch 847/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 744.6596 - mean_absolute_error: 16.5493 - mean_absolute_percentage_error: 16.1460\n", "Epoch 848/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 744.2991 - mean_absolute_error: 16.5454 - mean_absolute_percentage_error: 16.1435\n", "Epoch 849/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 743.9385 - mean_absolute_error: 16.5415 - mean_absolute_percentage_error: 16.1410\n", "Epoch 850/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 743.5781 - mean_absolute_error: 16.5378 - mean_absolute_percentage_error: 16.1389\n", "Epoch 851/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 743.2181 - mean_absolute_error: 16.5342 - mean_absolute_percentage_error: 16.1369\n", "Epoch 852/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 742.8584 - mean_absolute_error: 16.5306 - mean_absolute_percentage_error: 16.1350\n", "Epoch 853/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 742.4994 - mean_absolute_error: 16.5271 - mean_absolute_percentage_error: 16.1331\n", "Epoch 854/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 742.1412 - mean_absolute_error: 16.5236 - mean_absolute_percentage_error: 16.1312\n", "Epoch 855/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 741.7838 - mean_absolute_error: 16.5202 - mean_absolute_percentage_error: 16.1294\n", "Epoch 856/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 741.4274 - mean_absolute_error: 16.5169 - mean_absolute_percentage_error: 16.1279\n", "Epoch 857/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 741.0724 - mean_absolute_error: 16.5137 - mean_absolute_percentage_error: 16.1264\n", "Epoch 858/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 740.7188 - mean_absolute_error: 16.5105 - mean_absolute_percentage_error: 16.1250\n", "Epoch 859/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 740.3665 - mean_absolute_error: 16.5074 - mean_absolute_percentage_error: 16.1237\n", "Epoch 860/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 740.0157 - mean_absolute_error: 16.5043 - mean_absolute_percentage_error: 16.1225\n", "Epoch 861/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 739.6671 - mean_absolute_error: 16.5012 - mean_absolute_percentage_error: 16.1212\n", "Epoch 862/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 739.3200 - mean_absolute_error: 16.4982 - mean_absolute_percentage_error: 16.1199\n", "Epoch 863/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 738.9750 - mean_absolute_error: 16.4951 - mean_absolute_percentage_error: 16.1187\n", "Epoch 864/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 738.6321 - mean_absolute_error: 16.4921 - mean_absolute_percentage_error: 16.1174\n", "Epoch 865/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 738.2913 - mean_absolute_error: 16.4891 - mean_absolute_percentage_error: 16.1161\n", "Epoch 866/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 737.9527 - mean_absolute_error: 16.4862 - mean_absolute_percentage_error: 16.1150\n", "Epoch 867/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 737.6165 - mean_absolute_error: 16.4834 - mean_absolute_percentage_error: 16.1138\n", "Epoch 868/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 737.2825 - mean_absolute_error: 16.4805 - mean_absolute_percentage_error: 16.1126\n", "Epoch 869/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 736.9508 - mean_absolute_error: 16.4777 - mean_absolute_percentage_error: 16.1115\n", "Epoch 870/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 736.6217 - mean_absolute_error: 16.4748 - mean_absolute_percentage_error: 16.1103\n", "Epoch 871/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 736.2949 - mean_absolute_error: 16.4720 - mean_absolute_percentage_error: 16.1091\n", "Epoch 872/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 735.9706 - mean_absolute_error: 16.4692 - mean_absolute_percentage_error: 16.1079\n", "Epoch 873/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 735.6488 - mean_absolute_error: 16.4664 - mean_absolute_percentage_error: 16.1066\n", "Epoch 874/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 735.3295 - mean_absolute_error: 16.4636 - mean_absolute_percentage_error: 16.1054\n", "Epoch 875/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 735.0126 - mean_absolute_error: 16.4608 - mean_absolute_percentage_error: 16.1042\n", "Epoch 876/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 734.6980 - mean_absolute_error: 16.4580 - mean_absolute_percentage_error: 16.1029\n", "Epoch 877/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 734.3860 - mean_absolute_error: 16.4553 - mean_absolute_percentage_error: 16.1016\n", "Epoch 878/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 734.0763 - mean_absolute_error: 16.4525 - mean_absolute_percentage_error: 16.1004\n", "Epoch 879/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 733.7690 - mean_absolute_error: 16.4498 - mean_absolute_percentage_error: 16.0992\n", "Epoch 880/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 733.4640 - mean_absolute_error: 16.4473 - mean_absolute_percentage_error: 16.0981\n", "Epoch 881/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 733.1612 - mean_absolute_error: 16.4448 - mean_absolute_percentage_error: 16.0971\n", "Epoch 882/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 732.8607 - mean_absolute_error: 16.4423 - mean_absolute_percentage_error: 16.0960\n", "Epoch 883/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 732.5622 - mean_absolute_error: 16.4398 - mean_absolute_percentage_error: 16.0949\n", "Epoch 884/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 732.2657 - mean_absolute_error: 16.4373 - mean_absolute_percentage_error: 16.0939\n", "Epoch 885/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 731.9714 - mean_absolute_error: 16.4349 - mean_absolute_percentage_error: 16.0928\n", "Epoch 886/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 731.6790 - mean_absolute_error: 16.4325 - mean_absolute_percentage_error: 16.0918\n", "Epoch 887/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 731.3883 - mean_absolute_error: 16.4301 - mean_absolute_percentage_error: 16.0908\n", "Epoch 888/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 731.0995 - mean_absolute_error: 16.4278 - mean_absolute_percentage_error: 16.0898\n", "Epoch 889/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 730.8125 - mean_absolute_error: 16.4254 - mean_absolute_percentage_error: 16.0887\n", "Epoch 890/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 730.5270 - mean_absolute_error: 16.4230 - mean_absolute_percentage_error: 16.0877\n", "Epoch 891/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 730.2432 - mean_absolute_error: 16.4206 - mean_absolute_percentage_error: 16.0867\n", "Epoch 892/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 729.9609 - mean_absolute_error: 16.4183 - mean_absolute_percentage_error: 16.0856\n", "Epoch 893/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 729.6800 - mean_absolute_error: 16.4159 - mean_absolute_percentage_error: 16.0846\n", "Epoch 894/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 729.4006 - mean_absolute_error: 16.4136 - mean_absolute_percentage_error: 16.0835\n", "Epoch 895/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 729.1224 - mean_absolute_error: 16.4112 - mean_absolute_percentage_error: 16.0825\n", "Epoch 896/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 728.8458 - mean_absolute_error: 16.4089 - mean_absolute_percentage_error: 16.0814\n", "Epoch 897/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 728.5703 - mean_absolute_error: 16.4066 - mean_absolute_percentage_error: 16.0804\n", "Epoch 898/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 728.2960 - mean_absolute_error: 16.4042 - mean_absolute_percentage_error: 16.0793\n", "Epoch 899/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 728.0228 - mean_absolute_error: 16.4019 - mean_absolute_percentage_error: 16.0783\n", "Epoch 900/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 727.7509 - mean_absolute_error: 16.3996 - mean_absolute_percentage_error: 16.0772\n", "Epoch 901/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 727.4799 - mean_absolute_error: 16.3973 - mean_absolute_percentage_error: 16.0762\n", "Epoch 902/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 727.2100 - mean_absolute_error: 16.3950 - mean_absolute_percentage_error: 16.0751\n", "Epoch 903/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 726.9410 - mean_absolute_error: 16.3927 - mean_absolute_percentage_error: 16.0741\n", "Epoch 904/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 726.6732 - mean_absolute_error: 16.3904 - mean_absolute_percentage_error: 16.0730\n", "Epoch 905/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 726.4062 - mean_absolute_error: 16.3881 - mean_absolute_percentage_error: 16.0719\n", "Epoch 906/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 726.1401 - mean_absolute_error: 16.3858 - mean_absolute_percentage_error: 16.0709\n", "Epoch 907/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 725.8750 - mean_absolute_error: 16.3835 - mean_absolute_percentage_error: 16.0698\n", "Epoch 908/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 725.6106 - mean_absolute_error: 16.3813 - mean_absolute_percentage_error: 16.0688\n", "Epoch 909/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 725.3471 - mean_absolute_error: 16.3791 - mean_absolute_percentage_error: 16.0678\n", "Epoch 910/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 725.0844 - mean_absolute_error: 16.3769 - mean_absolute_percentage_error: 16.0668\n", "Epoch 911/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 724.8226 - mean_absolute_error: 16.3747 - mean_absolute_percentage_error: 16.0658\n", "Epoch 912/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 724.5615 - mean_absolute_error: 16.3725 - mean_absolute_percentage_error: 16.0648\n", "Epoch 913/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 724.3009 - mean_absolute_error: 16.3703 - mean_absolute_percentage_error: 16.0638\n", "Epoch 914/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 724.0413 - mean_absolute_error: 16.3681 - mean_absolute_percentage_error: 16.0628\n", "Epoch 915/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 723.7823 - mean_absolute_error: 16.3659 - mean_absolute_percentage_error: 16.0619\n", "Epoch 916/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 723.5239 - mean_absolute_error: 16.3638 - mean_absolute_percentage_error: 16.0609\n", "Epoch 917/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 723.2662 - mean_absolute_error: 16.3616 - mean_absolute_percentage_error: 16.0599\n", "Epoch 918/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 723.0091 - mean_absolute_error: 16.3595 - mean_absolute_percentage_error: 16.0589\n", "Epoch 919/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 722.7526 - mean_absolute_error: 16.3573 - mean_absolute_percentage_error: 16.0579\n", "Epoch 920/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 722.4968 - mean_absolute_error: 16.3552 - mean_absolute_percentage_error: 16.0569\n", "Epoch 921/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 722.2415 - mean_absolute_error: 16.3531 - mean_absolute_percentage_error: 16.0559\n", "Epoch 922/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 721.9868 - mean_absolute_error: 16.3509 - mean_absolute_percentage_error: 16.0550\n", "Epoch 923/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 721.7325 - mean_absolute_error: 16.3488 - mean_absolute_percentage_error: 16.0540\n", "Epoch 924/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 721.4788 - mean_absolute_error: 16.3467 - mean_absolute_percentage_error: 16.0530\n", "Epoch 925/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 721.2256 - mean_absolute_error: 16.3445 - mean_absolute_percentage_error: 16.0520\n", "Epoch 926/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 720.9730 - mean_absolute_error: 16.3424 - mean_absolute_percentage_error: 16.0511\n", "Epoch 927/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 720.7207 - mean_absolute_error: 16.3403 - mean_absolute_percentage_error: 16.0501\n", "Epoch 928/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 720.4688 - mean_absolute_error: 16.3382 - mean_absolute_percentage_error: 16.0492\n", "Epoch 929/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 720.2175 - mean_absolute_error: 16.3361 - mean_absolute_percentage_error: 16.0482\n", "Epoch 930/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 719.9665 - mean_absolute_error: 16.3339 - mean_absolute_percentage_error: 16.0473\n", "Epoch 931/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 719.7159 - mean_absolute_error: 16.3318 - mean_absolute_percentage_error: 16.0463\n", "Epoch 932/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 719.4657 - mean_absolute_error: 16.3297 - mean_absolute_percentage_error: 16.0453\n", "Epoch 933/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 719.2159 - mean_absolute_error: 16.3275 - mean_absolute_percentage_error: 16.0444\n", "Epoch 934/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 718.9664 - mean_absolute_error: 16.3254 - mean_absolute_percentage_error: 16.0434\n", "Epoch 935/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 718.7172 - mean_absolute_error: 16.3233 - mean_absolute_percentage_error: 16.0425\n", "Epoch 936/1000\n", "1/1 [==============================] - 0s 10ms/step - loss: 718.4682 - mean_absolute_error: 16.3212 - mean_absolute_percentage_error: 16.0416\n", "Epoch 937/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 718.2197 - mean_absolute_error: 16.3191 - mean_absolute_percentage_error: 16.0406\n", "Epoch 938/1000\n", "1/1 [==============================] - 0s 11ms/step - loss: 717.9713 - mean_absolute_error: 16.3169 - mean_absolute_percentage_error: 16.0397\n", "Epoch 939/1000\n", "1/1 [==============================] - 0s 13ms/step - loss: 717.7233 - mean_absolute_error: 16.3148 - mean_absolute_percentage_error: 16.0387\n", "Epoch 940/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 717.4755 - mean_absolute_error: 16.3127 - mean_absolute_percentage_error: 16.0378\n", "Epoch 941/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 717.2278 - mean_absolute_error: 16.3105 - mean_absolute_percentage_error: 16.0368\n", "Epoch 942/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 716.9803 - mean_absolute_error: 16.3084 - mean_absolute_percentage_error: 16.0359\n", "Epoch 943/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 716.7332 - mean_absolute_error: 16.3062 - mean_absolute_percentage_error: 16.0349\n", "Epoch 944/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 716.4861 - mean_absolute_error: 16.3040 - mean_absolute_percentage_error: 16.0339\n", "Epoch 945/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 716.2390 - mean_absolute_error: 16.3019 - mean_absolute_percentage_error: 16.0330\n", "Epoch 946/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 715.9922 - mean_absolute_error: 16.2997 - mean_absolute_percentage_error: 16.0320\n", "Epoch 947/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 715.7457 - mean_absolute_error: 16.2975 - mean_absolute_percentage_error: 16.0311\n", "Epoch 948/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 715.4990 - mean_absolute_error: 16.2954 - mean_absolute_percentage_error: 16.0301\n", "Epoch 949/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 715.2526 - mean_absolute_error: 16.2932 - mean_absolute_percentage_error: 16.0291\n", "Epoch 950/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 715.0062 - mean_absolute_error: 16.2910 - mean_absolute_percentage_error: 16.0282\n", "Epoch 951/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 714.7598 - mean_absolute_error: 16.2888 - mean_absolute_percentage_error: 16.0272\n", "Epoch 952/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 714.5134 - mean_absolute_error: 16.2866 - mean_absolute_percentage_error: 16.0262\n", "Epoch 953/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 714.2672 - mean_absolute_error: 16.2844 - mean_absolute_percentage_error: 16.0253\n", "Epoch 954/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 714.0209 - mean_absolute_error: 16.2822 - mean_absolute_percentage_error: 16.0243\n", "Epoch 955/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 713.7747 - mean_absolute_error: 16.2801 - mean_absolute_percentage_error: 16.0234\n", "Epoch 956/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 713.5284 - mean_absolute_error: 16.2779 - mean_absolute_percentage_error: 16.0224\n", "Epoch 957/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 713.2822 - mean_absolute_error: 16.2757 - mean_absolute_percentage_error: 16.0215\n", "Epoch 958/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 713.0359 - mean_absolute_error: 16.2736 - mean_absolute_percentage_error: 16.0206\n", "Epoch 959/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 712.7894 - mean_absolute_error: 16.2714 - mean_absolute_percentage_error: 16.0196\n", "Epoch 960/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 712.5431 - mean_absolute_error: 16.2692 - mean_absolute_percentage_error: 16.0186\n", "Epoch 961/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 712.2966 - mean_absolute_error: 16.2670 - mean_absolute_percentage_error: 16.0177\n", "Epoch 962/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 712.0501 - mean_absolute_error: 16.2649 - mean_absolute_percentage_error: 16.0168\n", "Epoch 963/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 711.8035 - mean_absolute_error: 16.2627 - mean_absolute_percentage_error: 16.0158\n", "Epoch 964/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 711.5568 - mean_absolute_error: 16.2605 - mean_absolute_percentage_error: 16.0149\n", "Epoch 965/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 711.3101 - mean_absolute_error: 16.2584 - mean_absolute_percentage_error: 16.0139\n", "Epoch 966/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 711.0632 - mean_absolute_error: 16.2562 - mean_absolute_percentage_error: 16.0130\n", "Epoch 967/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 710.8163 - mean_absolute_error: 16.2540 - mean_absolute_percentage_error: 16.0120\n", "Epoch 968/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 710.5691 - mean_absolute_error: 16.2519 - mean_absolute_percentage_error: 16.0111\n", "Epoch 969/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 710.3220 - mean_absolute_error: 16.2497 - mean_absolute_percentage_error: 16.0101\n", "Epoch 970/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 710.0746 - mean_absolute_error: 16.2475 - mean_absolute_percentage_error: 16.0092\n", "Epoch 971/1000\n", "1/1 [==============================] - 0s 10ms/step - loss: 709.8273 - mean_absolute_error: 16.2454 - mean_absolute_percentage_error: 16.0083\n", "Epoch 972/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 709.5798 - mean_absolute_error: 16.2432 - mean_absolute_percentage_error: 16.0073\n", "Epoch 973/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 709.3322 - mean_absolute_error: 16.2410 - mean_absolute_percentage_error: 16.0064\n", "Epoch 974/1000\n", "1/1 [==============================] - 0s 6ms/step - loss: 709.0845 - mean_absolute_error: 16.2388 - mean_absolute_percentage_error: 16.0055\n", "Epoch 975/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 708.8367 - mean_absolute_error: 16.2366 - mean_absolute_percentage_error: 16.0045\n", "Epoch 976/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 708.5887 - mean_absolute_error: 16.2344 - mean_absolute_percentage_error: 16.0036\n", "Epoch 977/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 708.3407 - mean_absolute_error: 16.2322 - mean_absolute_percentage_error: 16.0026\n", "Epoch 978/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 708.0924 - mean_absolute_error: 16.2299 - mean_absolute_percentage_error: 16.0017\n", "Epoch 979/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 707.8441 - mean_absolute_error: 16.2277 - mean_absolute_percentage_error: 16.0007\n", "Epoch 980/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 707.5957 - mean_absolute_error: 16.2255 - mean_absolute_percentage_error: 15.9998\n", "Epoch 981/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 707.3472 - mean_absolute_error: 16.2232 - mean_absolute_percentage_error: 15.9989\n", "Epoch 982/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 707.0986 - mean_absolute_error: 16.2210 - mean_absolute_percentage_error: 15.9979\n", "Epoch 983/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 706.8497 - mean_absolute_error: 16.2187 - mean_absolute_percentage_error: 15.9970\n", "Epoch 984/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 706.6008 - mean_absolute_error: 16.2165 - mean_absolute_percentage_error: 15.9960\n", "Epoch 985/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 706.3518 - mean_absolute_error: 16.2142 - mean_absolute_percentage_error: 15.9950\n", "Epoch 986/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 706.1026 - mean_absolute_error: 16.2119 - mean_absolute_percentage_error: 15.9941\n", "Epoch 987/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 705.8533 - mean_absolute_error: 16.2096 - mean_absolute_percentage_error: 15.9931\n", "Epoch 988/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 705.6038 - mean_absolute_error: 16.2074 - mean_absolute_percentage_error: 15.9922\n", "Epoch 989/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 705.3541 - mean_absolute_error: 16.2051 - mean_absolute_percentage_error: 15.9912\n", "Epoch 990/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 705.1041 - mean_absolute_error: 16.2028 - mean_absolute_percentage_error: 15.9903\n", "Epoch 991/1000\n", "1/1 [==============================] - 0s 9ms/step - loss: 704.8539 - mean_absolute_error: 16.2005 - mean_absolute_percentage_error: 15.9894\n", "Epoch 992/1000\n", "1/1 [==============================] - 0s 4ms/step - loss: 704.6035 - mean_absolute_error: 16.1982 - mean_absolute_percentage_error: 15.9885\n", "Epoch 993/1000\n", "1/1 [==============================] - 0s 7ms/step - loss: 704.3528 - mean_absolute_error: 16.1960 - mean_absolute_percentage_error: 15.9876\n", "Epoch 994/1000\n", "1/1 [==============================] - 0s 5ms/step - loss: 704.1017 - mean_absolute_error: 16.1937 - mean_absolute_percentage_error: 15.9868\n", "Epoch 995/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 703.8502 - mean_absolute_error: 16.1915 - mean_absolute_percentage_error: 15.9859\n", "Epoch 996/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 703.5984 - mean_absolute_error: 16.1893 - mean_absolute_percentage_error: 15.9851\n", "Epoch 997/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 703.3460 - mean_absolute_error: 16.1871 - mean_absolute_percentage_error: 15.9843\n", "Epoch 998/1000\n", "1/1 [==============================] - 0s 3ms/step - loss: 703.0931 - mean_absolute_error: 16.1849 - mean_absolute_percentage_error: 15.9836\n", "Epoch 999/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 702.8395 - mean_absolute_error: 16.1827 - mean_absolute_percentage_error: 15.9829\n", "Epoch 1000/1000\n", "1/1 [==============================] - 0s 2ms/step - loss: 702.5853 - mean_absolute_error: 16.1805 - mean_absolute_percentage_error: 15.9821\n" ], "name": "stdout" } ] }, { "cell_type": "code", "metadata": { "id": "dJiTro25nYVS", "outputId": "b562c25b-8f9a-4e79-bdd5-80f557defd19", "colab": { "base_uri": "https://localhost:8080/", "height": 54 } }, "source": [ "model2.evaluate(x_test_norm, y_test)" ], "execution_count": 27, "outputs": [ { "output_type": "stream", "text": [ "4/4 [==============================] - 0s 2ms/step - loss: 1324.2524 - mean_absolute_error: 21.4850 - mean_absolute_percentage_error: 20.1677\n" ], "name": "stdout" }, { "output_type": "execute_result", "data": { "text/plain": [ "[1324.25244140625, 21.485002517700195, 20.16765594482422]" ] }, "metadata": { "tags": [] }, "execution_count": 27 } ] }, { "cell_type": "code", "metadata": { "id": "p9DGbxa8na-9", "outputId": "4a7fce96-ff60-4adf-d3b7-29817e26b7c6", "colab": { "base_uri": "https://localhost:8080/", "height": 328 } }, "source": [ "plt.plot(history_with_minibatch.history['loss'], label='With minibatch',linewidth=1.5)\n", "plt.title('Training Mean Squared Error (MSE)\\nwith neural network architecture (10-20-2)', fontsize=12)\n", "plt.xlabel('epochs')\n", "plt.ylabel('Loss')\n", "plt.plot(history_without_minibatch.history['loss'], label='Without minibatch',linewidth=1.5)\n", "plt.legend()\n", "plt.xlim([-10, 1000])\n", "#plt.plot(test_history_without_minibatch.losses.T[0])" ], "execution_count": 28, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "(-10.0, 1000.0)" ] }, "metadata": { "tags": [] }, "execution_count": 28 }, { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [], "needs_background": "light" } } ] }, { "cell_type": "markdown", "metadata": { "id": "eykmk_nXne_g" }, "source": [ "As can be seen above, with the use of minibatches, the training converged much faster." ] }, { "cell_type": "markdown", "metadata": { "id": "A2-kQZw4nwgS" }, "source": [ "### Changing number of neurons in the hidden layer\n", "\n", "Up to this point, the number os neurons in the hidden layer was chosen rather arbitrarily as 20. To experiment, with this, the architectures of (10−10−2), (10−20−2), (10−30−2), (10−40−2), and (10−50−2) will be considered, keeping the rest as before." ] }, { "cell_type": "code", "metadata": { "id": "33bvqFGNneeG" }, "source": [ "# Creates a model with the specific number of neurons\n", "def make_model_0(num_neurons):\n", " model = models.Sequential()\n", " model.add(layers.Dense(units=num_neurons, activation='sigmoid', input_shape=(10,)))\n", " model.add(layers.Dense(2))\n", "\n", " model.compile(optimizer=sgd,\n", " loss='mean_squared_error',\n", " metrics=['mean_absolute_error', 'mean_absolute_percentage_error'])\n", " return model" ], "execution_count": 30, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "McBJqYXF3lDa" }, "source": [ "model_10_neurons = make_model_0(10)\n", "model_20_neurons = make_model_0(20)\n", "model_30_neurons = make_model_0(30)\n", "model_40_neurons = make_model_0(40)\n", "model_50_neurons = make_model_0(50)" ], "execution_count": 31, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "TFN--Tlin3zu", "outputId": "c37d6488-fe2b-49aa-b373-f329720d0f0e", "colab": { "base_uri": "https://localhost:8080/", "height": 123 } }, "source": [ "# Training the models - output will be suppressed\n", "print('10 neurons')\n", "hist_10_neurons = model_10_neurons.fit(x_train_norm, y_train, epochs=1000, verbose=0)\n", "print('20 neurons')\n", "hist_20_neurons = model_20_neurons.fit(x_train_norm, y_train, epochs=1000, verbose=0)\n", "print('30 neurons')\n", "hist_30_neurons = model_30_neurons.fit(x_train_norm, y_train, epochs=1000, verbose=0)\n", "print('40 neurons')\n", "hist_40_neurons = model_40_neurons.fit(x_train_norm, y_train, epochs=1000, verbose=0)\n", "print('50 neurons')\n", "hist_50_neurons = model_50_neurons.fit(x_train_norm, y_train, epochs=1000, verbose=0)\n", "print('Done!')" ], "execution_count": null, "outputs": [ { "output_type": "stream", "text": [ "10 neurons\n", "20 neurons\n", "30 neurons\n", "40 neurons\n", "50 neurons\n", "Done!\n" ], "name": "stdout" } ] }, { "cell_type": "code", "metadata": { "id": "ed8Cl5j7n50I", "outputId": "327eb57c-f125-4dde-a06c-013b2d449768", "colab": { "base_uri": "https://localhost:8080/", "height": 310 } }, "source": [ "plt.title('MSE for the models with\\nvarying number of neurons in hidden layers', fontsize=12)\n", "plt.xlabel('epochs')\n", "plt.ylabel('Loss')\n", "plt.plot(hist_10_neurons.history['loss'], label='10',linewidth=1.0)\n", "plt.plot(hist_20_neurons.history['loss'], label='20',linewidth=1.0)\n", "plt.plot(hist_30_neurons.history['loss'], label='30',linewidth=1.0)\n", "plt.plot(hist_40_neurons.history['loss'], label='40',linewidth=1.0)\n", "plt.plot(hist_50_neurons.history['loss'], label='50',linewidth=1.0)\n", "plt.ylim([0,2500])\n", "plt.legend();" ], "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [], "needs_background": "light" } } ] }, { "cell_type": "markdown", "metadata": { "id": "0sAmn-hyn8vQ" }, "source": [ "From the plot above, it can be seen that the learning gets faster as the number of neurons increases. However, from 40 to 50 neurons the difference is less significative than from 10 to 20 neurons." ] }, { "cell_type": "markdown", "metadata": { "id": "_w7CihwB6d7j" }, "source": [ "### Cross validation\n", "\n", "When the amount of input data is large enough, one can divide the data into *training, validation* and *test* set. The *training set*, as the name suggests, is used to train the model, that is, to adjust the internal parameters of the model such that the inputs match the outputs with the minimum error possible. The *validation set* is used to test the performance of the model before it is subject to the actual test set, and to tune its hyperparameters. The *test set*, which should be kept separate and untouched in the training process, is used to provide an unbiased evaluation of the performance of the final model.\n", "\n", " ![](https://drive.google.com/uc?export=view&id=19txDsk2QD64Y2zjh5c439e8U5KM7j_UZ)\n", "\n", "\n", "When there is no validation set, the most widely used approach is cross-validation (CV). In the well known K-fold CV the training set is split into $K$ parts - see figure below. The input data set is randomly divided into $K$ subsets (also known as folds). The ML model is trained with $K-1$ subsets, and evaluated in the subset that was not used for training. This process is repeated $K$ times with a different subset reserved for evaluation (and excluded from training) each time. For each training step, the average error is calculated. At the end, the model with the smallest error is selected. This can simulate training/validation, which is useful for hyperparameter tuning, without touching the test set. \n", "\n", " ![](https://drive.google.com/uc?export=view&id=15k96E0QY-QtKpELudeqMMpsApCxibAHJ)\n", "\n" ] }, { "cell_type": "markdown", "metadata": { "id": "vW-kekXooxAk" }, "source": [ "Applying cross-validation, seeking for the best model:" ] }, { "cell_type": "code", "metadata": { "id": "83p91hhQowE_" }, "source": [ "from sklearn.model_selection import cross_val_score, KFold" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "Yvoc-7Zzo_HE" }, "source": [ "def get_cross_val_score(model, x, y, cv=5, epochs=1000):\n", " cvs = np.zeros((cv, len(model.metrics)))\n", " \n", " k_folds = KFold(n_splits=cv)\n", " k_folds.split(x, y)\n", " for j, (train_idx, test_idx) in enumerate(k_folds.split(x, y)):\n", " model.fit(x_train_norm[train_idx], y_train[train_idx], epochs=100, verbose=0)\n", " cvs[j,:] = np.array(model.evaluate(x=x_train_norm[test_idx], y=y_train[test_idx], verbose=0))\n", " return cvs" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "cPmqySCNpApW", "outputId": "e6314791-5284-451a-a5c4-4abf941531b4", "colab": { "base_uri": "https://localhost:8080/", "height": 123 } }, "source": [ "print('10 neurons')\n", "cv_10_neurons = get_cross_val_score(model_10_neurons, x_train_norm, y_train, cv=5, epochs=1000)\n", "print('20 neurons')\n", "cv_20_neurons = get_cross_val_score(model_20_neurons, x_train_norm, y_train, cv=5, epochs=1000)\n", "print('30 neurons')\n", "cv_30_neurons = get_cross_val_score(model_30_neurons, x_train_norm, y_train, cv=5, epochs=1000)\n", "print('40 neurons')\n", "cv_40_neurons = get_cross_val_score(model_40_neurons, x_train_norm, y_train, cv=5, epochs=1000)\n", "print('50 neurons')\n", "cv_50_neurons = get_cross_val_score(model_50_neurons, x_train_norm, y_train, cv=5, epochs=1000)\n", "\n", "print('Done')" ], "execution_count": null, "outputs": [ { "output_type": "stream", "text": [ "10 neurons\n", "20 neurons\n", "30 neurons\n", "40 neurons\n", "50 neurons\n", "Done\n" ], "name": "stdout" } ] }, { "cell_type": "code", "metadata": { "id": "_2w8uisFpDJ0", "outputId": "b60e9718-a863-4184-ab04-e4c35898dc9f", "colab": { "base_uri": "https://localhost:8080/", "height": 312 } }, "source": [ "plt.style.use('bmh')\n", "fig, ax = plt.subplots()\n", "\n", "xi = np.arange(10,60,10)\n", "means = np.array([cv_10_neurons[:,0].mean(),\n", " cv_20_neurons[:,0].mean(),\n", " cv_30_neurons[:,0].mean(),\n", " cv_40_neurons[:,0].mean(),\n", " cv_50_neurons[:,0].mean()])\n", "\n", "stds = np.array([cv_10_neurons[:,0].std(),\n", " cv_20_neurons[:,0].std(),\n", " cv_30_neurons[:,0].std(),\n", " cv_40_neurons[:,0].std(),\n", " cv_50_neurons[:,0].std()])\n", "\n", "ax.errorbar(xi,\n", " means,\n", " yerr=stds);\n", "ax.set(xticks=(xi),\n", " title='MSE\\'s mean plus or minus one standard deviation\\nusing 5-fold cross-validation for different numbers of neurons',\n", " xlabel='Number of neurons',\n", " ylabel='MSE');" ], "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [], "needs_background": "light" } } ] }, { "cell_type": "code", "metadata": { "id": "E8wu7K0epGyY", "outputId": "254b36cb-3cac-48d2-c575-9ba910d45d4a", "colab": { "base_uri": "https://localhost:8080/", "height": 194 } }, "source": [ "predictions = np.zeros((5,3))\n", "predictions[0,:] = model_10_neurons.evaluate(x_test_norm, y_test)\n", "predictions[1,:] = model_20_neurons.evaluate(x_test_norm, y_test)\n", "predictions[2,:] = model_30_neurons.evaluate(x_test_norm, y_test)\n", "predictions[3,:] = model_40_neurons.evaluate(x_test_norm, y_test)\n", "predictions[4,:] = model_50_neurons.evaluate(x_test_norm, y_test)\n", "predictions" ], "execution_count": null, "outputs": [ { "output_type": "stream", "text": [ "4/4 [==============================] - 0s 2ms/step - loss: 1538.0431 - mean_absolute_error: 25.6651 - mean_absolute_percentage_error: 26.6588\n", "4/4 [==============================] - 0s 2ms/step - loss: 1753.5770 - mean_absolute_error: 29.7437 - mean_absolute_percentage_error: 31.8624\n", "4/4 [==============================] - 0s 2ms/step - loss: 1808.4327 - mean_absolute_error: 27.6502 - mean_absolute_percentage_error: 28.6878\n", "4/4 [==============================] - 0s 2ms/step - loss: 1527.5784 - mean_absolute_error: 23.8422 - mean_absolute_percentage_error: 24.7856\n", "4/4 [==============================] - 0s 2ms/step - loss: 1658.4849 - mean_absolute_error: 24.1849 - mean_absolute_percentage_error: 24.3957\n" ], "name": "stdout" }, { "output_type": "execute_result", "data": { "text/plain": [ "array([[1538.04309082, 25.66506195, 26.65881157],\n", " [1753.57702637, 29.74365997, 31.86240959],\n", " [1808.43273926, 27.65017891, 28.68782425],\n", " [1527.57836914, 23.84220314, 24.78555489],\n", " [1658.48486328, 24.18493843, 24.39570427]])" ] }, "metadata": { "tags": [] }, "execution_count": 47 } ] }, { "cell_type": "code", "metadata": { "id": "h_ASEb1CpIjj", "outputId": "0d97b7b6-bab9-4e1b-9d96-9ebe2a8e57f5", "colab": { "base_uri": "https://localhost:8080/", "height": 312 } }, "source": [ "fig, ax = plt.subplots()\n", "\n", "xi = np.arange(10,60,10)\n", "ax.plot(xi, predictions[:,0])\n", "ax.set(xticks=(xi),\n", " title='MSE for the test set\\nfor different numbers of neurons',\n", " xlabel='Number of neurons',\n", " ylabel='MSE');" ], "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [], "needs_background": "light" } } ] }, { "cell_type": "markdown", "metadata": { "id": "T7TtDgxQpVIu" }, "source": [ "## Activation functions\n", "\n", "Until now, the sigmoid or logistic function,\n", "$$g(z) = \\frac{1}{1+e^{-z}}$$\n", "was used to create activate the neurons.\n", "\n", "The purpose of the activation function is to introduce non-linearity into the output of a neuron. See the figure below for the most commonly used activation functions.\n", "\n", "![](https://drive.google.com/uc?export=view&id=1iZPDL1JuFLeH9uOc-SfbTfBWBnXwamRM)\n", "\n", "This non-linearity is one of the factors that affect our results and the accuracy of our model. When the NN has several hidden layers, a linear activation function will simply generate a series of related transformations so, this model is no more expressive than a simple standard logistic regression model. Unless we convey nonlinearity, we are not computing interesting models, even if we delve into neural networks.\n", "\n", "Some of these functions are plotted below." ] }, { "cell_type": "code", "metadata": { "id": "EsFGHTNtpaMu", "outputId": "d9e72a5f-0ca5-4d12-8f4d-d15ef32ab3c3", "colab": { "base_uri": "https://localhost:8080/", "height": 295 } }, "source": [ "x = np.linspace(-3, 3, 500)\n", "relu = np.maximum(0, x)\n", "sigmoid = 1/(1+np.exp(-x))\n", "tanh = np.tanh(x)\n", "softplus = np.log(1+np.exp(x))\n", "\n", "plt.plot(x, relu, label='ReLU',linewidth=1.5)\n", "plt.plot(x, sigmoid, label='sigmoid',linewidth=1.5)\n", "plt.plot(x, tanh, label='tanh',linewidth=1.5)\n", "plt.plot(x, softplus, label='softplus',linewidth=1.5)\n", "plt.legend(fancybox=True, framealpha=0.5, loc='upper right')\n", "plt.title('Some common activation functions', fontsize=12)" ], "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "Text(0.5, 1.0, 'Some common activation functions')" ] }, "metadata": { "tags": [] }, "execution_count": 54 }, { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [], "needs_background": "light" } } ] }, { "cell_type": "markdown", "metadata": { "id": "kTQIG5c0phTF" }, "source": [ "### Training again with other activations\n", "\n", "Now, other activations will be used in the model and its result will be evaluated. The functions that will be used are sigmoid, ReLU, tanh e softplus. Other parameters and hyperparameters will be kept the same from the previous model, with SGD as optimizer, MSE as loss and with (10-20-2) architecture." ] }, { "cell_type": "code", "metadata": { "id": "PcXbHtQipjTy" }, "source": [ "# Creates a model changing only the activation function\n", "def make_model(activation):\n", " model = models.Sequential()\n", " model.add(layers.Dense(20, activation=activation, input_shape=(10,)))\n", " model.add(layers.Dense(2))\n", "\n", " model.compile(optimizer=sgd,\n", " loss='mean_squared_error',\n", " metrics=['mean_absolute_error', 'mean_absolute_percentage_error'])\n", " return model\n", "\n", "sigmoid_model = make_model('sigmoid')\n", "relu_model = make_model('relu')\n", "tanh_model = make_model('tanh')\n", "softplus_model = make_model('softplus')" ], "execution_count": 32, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "kBypJ2NRples" }, "source": [ "# Training the models\n", "sigmoid_history = sigmoid_model.fit(x_train_norm, y_train, epochs=1000, batch_size=32)\n", "relu_history = relu_model.fit(x_train_norm, y_train, epochs=1000, batch_size=32)\n", "tanh_history = tanh_model.fit(x_train_norm, y_train, epochs=1000, batch_size=32)\n", "softplus_history = softplus_model.fit(x_train_norm, y_train, epochs=1000, batch_size=32)" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "2ysGdmYKppJU", "outputId": "281cefa3-ab42-47c0-a478-d7d528b63f46", "colab": { "base_uri": "https://localhost:8080/", "height": 320 } }, "source": [ "plt.style.use('fivethirtyeight')\n", "plt.title('Training MSE for the models with different activation functions', fontsize=12)\n", "plt.xlabel('epochs')\n", "plt.ylabel('Loss')\n", "plt.plot(sigmoid_history.history['loss'], label='sigmoid',linewidth=1.5)\n", "plt.plot(relu_history.history['loss'], label='relu',linewidth=1.5)\n", "plt.plot(tanh_history.history['loss'], label='tanh',linewidth=1.5)\n", "plt.plot(softplus_history.history['loss'], label='softplus',linewidth=1.5)\n", "plt.ylim([0,2500])\n", "plt.legend(fancybox=True, framealpha=0.5)" ], "execution_count": null, "outputs": [ { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [] } } ] }, { "cell_type": "markdown", "metadata": { "id": "jEB29sNkqDxY" }, "source": [ "## Using all displacement data and also stress data\n", "\n", "As a reminder, the head for the dataframe featuring all the columns of the dataset is show below." ] }, { "cell_type": "code", "metadata": { "id": "DYyspWQZqFj2", "outputId": "307b4b57-83c1-4f50-8e2b-097671f60c48", "colab": { "base_uri": "https://localhost:8080/", "height": 247 } }, "source": [ "train.head()" ], "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
area1area2area3area4area5area6area7area8area9area10d1d2d3d4d5d6d7d8s11_1s11_2s11_3s11_4s11_5s11_6s11_7s11_8s11_9s11_10mises_1mises_2mises_3mises_4mises_5mises_6mises_7mises_8mises_9mises_10
iteration
2010.0209590.0052350.0191970.0100050.0200770.0023490.0012710.0106060.0210140.01053916.004036-120.325661-53.915527-120.18594416.137526-32.663422-50.967178-54.107288-22231914.01.616967e+08-1.006564e+06-1.053551e+06-384316160.0121684416.02838388.5011838953.0-62306460.065863956.022231914.01.616967e+081.006564e+061.053551e+06384316160.0121684416.02838388.5011838953.062306460.065863956.0
920.0225290.0058680.0116310.0167140.0156290.0098220.0175580.0110000.0161720.01112612.547656-53.592022-8.100519-54.67098211.504974-27.936739-6.230652-30.526878-14099667.01.953085e+077.862298e+068.135870e+06-46982004.086752840.0-16162369.0091602400.0-61951560.026877044.014099667.01.953085e+077.862298e+068.135870e+0646982004.086752840.016162369.0091602400.061951560.026877044.0
3440.0032190.0123110.0178840.0161570.0032170.0211460.0210480.0081380.0220800.00435014.224157-54.255360-18.496674-60.12879613.368400-21.245007-6.041500-21.704147-93917784.03.462124e+066.452812e+064.428844e+07-45555708.0100803928.0-46319392.0059051812.0-29696636.026462190.093917784.03.462124e+066.452812e+064.428844e+0745555708.0100803928.046319392.0059051812.029696636.026462190.0
1190.0152170.0201340.0039380.0145350.0110370.0077390.0160590.0220210.0005480.02026718.363760-50.452175-12.587241-51.1659433.985027-24.647602-9.228423-21.760574-25327046.0-2.176952e+071.084223e+085.382156e+06-69586584.030048950.0-4145055.7547249120.0-77902696.037499008.025327046.02.176952e+071.084223e+085.382156e+0669586584.030048950.04145055.7547249120.077902696.037499008.0
2210.0004900.0029860.0065510.0106980.0006630.0015720.0179950.0221350.0097660.0106408.433659-77.570465-70.816208-118.8271415.633955-34.009995-6.182016-55.686005-487371776.01.634472e+082.111106e+073.110945e+08-46615268.042482636.0-27404998.00186641504.0-106984240.031545470.0487371776.01.634472e+082.111106e+073.110945e+0846615268.042482636.027404998.00186641504.0106984240.031545470.0
\n", "
" ], "text/plain": [ " area1 area2 area3 area4 area5 area6 \\\n", "iteration \n", "201 0.020959 0.005235 0.019197 0.010005 0.020077 0.002349 \n", "92 0.022529 0.005868 0.011631 0.016714 0.015629 0.009822 \n", "344 0.003219 0.012311 0.017884 0.016157 0.003217 0.021146 \n", "119 0.015217 0.020134 0.003938 0.014535 0.011037 0.007739 \n", "221 0.000490 0.002986 0.006551 0.010698 0.000663 0.001572 \n", "\n", " area7 area8 area9 area10 d1 d2 \\\n", "iteration \n", "201 0.001271 0.010606 0.021014 0.010539 16.004036 -120.325661 \n", "92 0.017558 0.011000 0.016172 0.011126 12.547656 -53.592022 \n", "344 0.021048 0.008138 0.022080 0.004350 14.224157 -54.255360 \n", "119 0.016059 0.022021 0.000548 0.020267 18.363760 -50.452175 \n", "221 0.017995 0.022135 0.009766 0.010640 8.433659 -77.570465 \n", "\n", " d3 d4 d5 d6 d7 d8 \\\n", "iteration \n", "201 -53.915527 -120.185944 16.137526 -32.663422 -50.967178 -54.107288 \n", "92 -8.100519 -54.670982 11.504974 -27.936739 -6.230652 -30.526878 \n", "344 -18.496674 -60.128796 13.368400 -21.245007 -6.041500 -21.704147 \n", "119 -12.587241 -51.165943 3.985027 -24.647602 -9.228423 -21.760574 \n", "221 -70.816208 -118.827141 5.633955 -34.009995 -6.182016 -55.686005 \n", "\n", " s11_1 s11_2 s11_3 s11_4 s11_5 \\\n", "iteration \n", "201 -22231914.0 1.616967e+08 -1.006564e+06 -1.053551e+06 -384316160.0 \n", "92 -14099667.0 1.953085e+07 7.862298e+06 8.135870e+06 -46982004.0 \n", "344 -93917784.0 3.462124e+06 6.452812e+06 4.428844e+07 -45555708.0 \n", "119 -25327046.0 -2.176952e+07 1.084223e+08 5.382156e+06 -69586584.0 \n", "221 -487371776.0 1.634472e+08 2.111106e+07 3.110945e+08 -46615268.0 \n", "\n", " s11_6 s11_7 s11_8 s11_9 s11_10 \\\n", "iteration \n", "201 121684416.0 2838388.50 11838953.0 -62306460.0 65863956.0 \n", "92 86752840.0 -16162369.00 91602400.0 -61951560.0 26877044.0 \n", "344 100803928.0 -46319392.00 59051812.0 -29696636.0 26462190.0 \n", "119 30048950.0 -4145055.75 47249120.0 -77902696.0 37499008.0 \n", "221 42482636.0 -27404998.00 186641504.0 -106984240.0 31545470.0 \n", "\n", " mises_1 mises_2 mises_3 mises_4 mises_5 \\\n", "iteration \n", "201 22231914.0 1.616967e+08 1.006564e+06 1.053551e+06 384316160.0 \n", "92 14099667.0 1.953085e+07 7.862298e+06 8.135870e+06 46982004.0 \n", "344 93917784.0 3.462124e+06 6.452812e+06 4.428844e+07 45555708.0 \n", "119 25327046.0 2.176952e+07 1.084223e+08 5.382156e+06 69586584.0 \n", "221 487371776.0 1.634472e+08 2.111106e+07 3.110945e+08 46615268.0 \n", "\n", " mises_6 mises_7 mises_8 mises_9 mises_10 \n", "iteration \n", "201 121684416.0 2838388.50 11838953.0 62306460.0 65863956.0 \n", "92 86752840.0 16162369.00 91602400.0 61951560.0 26877044.0 \n", "344 100803928.0 46319392.00 59051812.0 29696636.0 26462190.0 \n", "119 30048950.0 4145055.75 47249120.0 77902696.0 37499008.0 \n", "221 42482636.0 27404998.00 186641504.0 106984240.0 31545470.0 " ] }, "metadata": { "tags": [] }, "execution_count": 59 } ] }, { "cell_type": "markdown", "metadata": { "id": "SNYDdEkEqIRg" }, "source": [ "The inputs will not change, but the outputs will. Now, $d_1$ to $d_8$ and also $S11_1$ to $S11_{10}$ will be used" ] }, { "cell_type": "code", "metadata": { "id": "JNvKW7GEqJuy", "outputId": "bb450e09-9fd7-43ed-f672-cee08a8c1d55", "colab": { "base_uri": "https://localhost:8080/", "height": 247 } }, "source": [ "new_y_train = train.loc[:, 'd1':'s11_10']\n", "new_y_test = test.loc[:, 'd1':'s11_10']\n", "new_y_train.head()" ], "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
d1d2d3d4d5d6d7d8s11_1s11_2s11_3s11_4s11_5s11_6s11_7s11_8s11_9s11_10
iteration
20116.004036-120.325661-53.915527-120.18594416.137526-32.663422-50.967178-54.107288-22231914.01.616967e+08-1.006564e+06-1.053551e+06-384316160.0121684416.02838388.5011838953.0-62306460.065863956.0
9212.547656-53.592022-8.100519-54.67098211.504974-27.936739-6.230652-30.526878-14099667.01.953085e+077.862298e+068.135870e+06-46982004.086752840.0-16162369.0091602400.0-61951560.026877044.0
34414.224157-54.255360-18.496674-60.12879613.368400-21.245007-6.041500-21.704147-93917784.03.462124e+066.452812e+064.428844e+07-45555708.0100803928.0-46319392.0059051812.0-29696636.026462190.0
11918.363760-50.452175-12.587241-51.1659433.985027-24.647602-9.228423-21.760574-25327046.0-2.176952e+071.084223e+085.382156e+06-69586584.030048950.0-4145055.7547249120.0-77902696.037499008.0
2218.433659-77.570465-70.816208-118.8271415.633955-34.009995-6.182016-55.686005-487371776.01.634472e+082.111106e+073.110945e+08-46615268.042482636.0-27404998.00186641504.0-106984240.031545470.0
\n", "
" ], "text/plain": [ " d1 d2 d3 d4 d5 d6 \\\n", "iteration \n", "201 16.004036 -120.325661 -53.915527 -120.185944 16.137526 -32.663422 \n", "92 12.547656 -53.592022 -8.100519 -54.670982 11.504974 -27.936739 \n", "344 14.224157 -54.255360 -18.496674 -60.128796 13.368400 -21.245007 \n", "119 18.363760 -50.452175 -12.587241 -51.165943 3.985027 -24.647602 \n", "221 8.433659 -77.570465 -70.816208 -118.827141 5.633955 -34.009995 \n", "\n", " d7 d8 s11_1 s11_2 s11_3 \\\n", "iteration \n", "201 -50.967178 -54.107288 -22231914.0 1.616967e+08 -1.006564e+06 \n", "92 -6.230652 -30.526878 -14099667.0 1.953085e+07 7.862298e+06 \n", "344 -6.041500 -21.704147 -93917784.0 3.462124e+06 6.452812e+06 \n", "119 -9.228423 -21.760574 -25327046.0 -2.176952e+07 1.084223e+08 \n", "221 -6.182016 -55.686005 -487371776.0 1.634472e+08 2.111106e+07 \n", "\n", " s11_4 s11_5 s11_6 s11_7 s11_8 \\\n", "iteration \n", "201 -1.053551e+06 -384316160.0 121684416.0 2838388.50 11838953.0 \n", "92 8.135870e+06 -46982004.0 86752840.0 -16162369.00 91602400.0 \n", "344 4.428844e+07 -45555708.0 100803928.0 -46319392.00 59051812.0 \n", "119 5.382156e+06 -69586584.0 30048950.0 -4145055.75 47249120.0 \n", "221 3.110945e+08 -46615268.0 42482636.0 -27404998.00 186641504.0 \n", "\n", " s11_9 s11_10 \n", "iteration \n", "201 -62306460.0 65863956.0 \n", "92 -61951560.0 26877044.0 \n", "344 -29696636.0 26462190.0 \n", "119 -77902696.0 37499008.0 \n", "221 -106984240.0 31545470.0 " ] }, "metadata": { "tags": [] }, "execution_count": 60 } ] }, { "cell_type": "markdown", "metadata": { "id": "td4qg2PwqMAA" }, "source": [ "Since the values of the displacements vary in many orders of magnitude from the stresses (~$10^2$ for displacements and ~$10^8$ for the stresses), the outputs will also be normalized." ] }, { "cell_type": "code", "metadata": { "id": "5gIW96bbqNy6" }, "source": [ "# Normalizing the output data using the normalizer from scikit-learn\n", "normalizer = preprocessing.StandardScaler().fit(new_y_train)\n", "y_train_norm = normalizer.transform(new_y_train)\n", "y_test_norm = normalizer.transform(new_y_test)" ], "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "MJr8jVESqQzS" }, "source": [ "A comparison with different optimizers and activation functions will be made. Models with (10-20-18) arquitecture will be made using these diffentent combinations and the results will be gathered and shown.\n", "\n", "The optimizers that will be used are:\n", "+ SGD\n", "+ AdaGrad\n", "+ Adadelta\n", "+ RMSprop\n", "+ Adam\n", "\n", "And the activation functions will be:\n", "+ sigmoid\n", "+ tanh\n", "+ softplus\n", "+ ReLU" ] }, { "cell_type": "code", "metadata": { "id": "R69b-_iHqTB5" }, "source": [ "# Makes a model with different optimizer and activation\n", "def make_model_2(activation, optimizer):\n", " model = models.Sequential()\n", " model.add(layers.Dense(20, activation=activation, input_shape=(10,)))\n", " model.add(layers.Dense(2))\n", "\n", " model.compile(optimizer=optimizer,\n", " loss='mean_squared_error',\n", " metrics=['mean_absolute_error', 'mean_absolute_percentage_error'])\n", " return model" ], "execution_count": 33, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "Dieth96YqUXs" }, "source": [ "models = {'SGD': {}, 'AdaGrad': {}, 'Adadelta': {}, 'RMSprop': {}, 'Adam': {}}\n", "activations = ['sigmoid', 'tanh', 'softplus', 'relu']" ], "execution_count": 34, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "WEGP9DiLqWNz", "outputId": "69b267ae-1b52-4730-9df1-86ee448d8149", "colab": { "base_uri": "https://localhost:8080/", "height": 372 } }, "source": [ "i = 0\n", "for optimizer in models.keys():\n", " for activation in activations:\n", " print(f'Iteration {i}: {optimizer} with {activation}')\n", " model = make_model_2(activation, optimizer)\n", " hist = model.fit(x_train_norm, y_train_norm, epochs=1000, verbose=0)\n", " train_loss = hist.history['loss']\n", " test_loss = model.evaluate(x_test_norm, y_test_norm, verbose=0)\n", " models[optimizer][activation] = {'model': model, 'train': train_loss, \n", " 'test': test_loss, 'hist': hist}\n", " i += 1" ], "execution_count": null, "outputs": [ { "output_type": "stream", "text": [ "Iteration 0: SGD with sigmoid\n", "Iteration 1: SGD with tanh\n", "Iteration 2: SGD with softplus\n", "Iteration 3: SGD with relu\n", "Iteration 4: AdaGrad with sigmoid\n", "Iteration 5: AdaGrad with tanh\n", "Iteration 6: AdaGrad with softplus\n", "Iteration 7: AdaGrad with relu\n", "Iteration 8: Adadelta with sigmoid\n", "Iteration 9: Adadelta with tanh\n", "Iteration 10: Adadelta with softplus\n", "Iteration 11: Adadelta with relu\n", "Iteration 12: RMSprop with sigmoid\n", "Iteration 13: RMSprop with tanh\n", "Iteration 14: RMSprop with softplus\n", "Iteration 15: RMSprop with relu\n", "Iteration 16: Adam with sigmoid\n", "Iteration 17: Adam with tanh\n", "Iteration 18: Adam with softplus\n", "Iteration 19: Adam with relu\n" ], "name": "stdout" } ] }, { "cell_type": "code", "metadata": { "id": "eY_TjbgjqX9R", "outputId": "e89fb77e-bcb8-4af9-efb5-9ff5f5b3d11f", "colab": { "base_uri": "https://localhost:8080/", "height": 352 } }, "source": [ "plt.style.use('fivethirtyeight')\n", "plt.title('Train MSE for diffentent optimizers\\nusing softplus as activation', fontsize=12)\n", "plt.xlabel('epochs')\n", "plt.ylabel('Loss')\n", "plt.plot(models['SGD']['softplus']['hist'].history['loss'], label='SGD',linewidth=1.5)\n", "plt.plot(models['AdaGrad']['softplus']['hist'].history['loss'], label='AdaGrad',linewidth=1.5)\n", "plt.plot(models['Adadelta']['softplus']['hist'].history['loss'], label='Adadelta',linewidth=1.5)\n", "plt.plot(models['RMSprop']['softplus']['hist'].history['loss'], label='RMSprop',linewidth=1.5)\n", "plt.plot(models['Adam']['softplus']['hist'].history['loss'], label='Adam',linewidth=1.5)\n", "plt.legend(fancybox=True, framealpha=0.5, fontsize=10)" ], "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "" ] }, "metadata": { "tags": [] }, "execution_count": 66 }, { "output_type": "display_data", "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "tags": [] } } ] }, { "cell_type": "markdown", "metadata": { "id": "9iyG6Pu3qakX" }, "source": [ "Using the softplus as activation function, it can be seen that the training is faster using the RMSprop and Adam optimizers." ] }, { "cell_type": "markdown", "metadata": { "id": "zsebV_dRqc6Z" }, "source": [ "To show the resulting test and training data in a dataframe, a pandas MultiIndex will be used. To create it, [the user guide from pandas](https://pandas.pydata.org/pandas-docs/stable/user_guide/advanced.html) was consulted." ] }, { "cell_type": "code", "metadata": { "id": "UkgWI0xnqfjV", "outputId": "08f8593f-66e8-4eab-bcee-82f18e4ac11e", "colab": { "base_uri": "https://localhost:8080/", "height": 212 } }, "source": [ "arrays = [['SGD', 'AdaGrad', 'Adadelta', 'RMSprop', 'Adam'],\n", " ['Training', 'Test']]\n", "index = pd.MultiIndex.from_product(arrays, names=['Optimizer', 'Train/Test'])\n", "index" ], "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "MultiIndex([( 'SGD', 'Training'),\n", " ( 'SGD', 'Test'),\n", " ( 'AdaGrad', 'Training'),\n", " ( 'AdaGrad', 'Test'),\n", " ('Adadelta', 'Training'),\n", " ('Adadelta', 'Test'),\n", " ( 'RMSprop', 'Training'),\n", " ( 'RMSprop', 'Test'),\n", " ( 'Adam', 'Training'),\n", " ( 'Adam', 'Test')],\n", " names=['Optimizer', 'Train/Test'])" ] }, "metadata": { "tags": [] }, "execution_count": 67 } ] }, { "cell_type": "code", "metadata": { "id": "Hv_ZKlyvqiCV", "outputId": "e4e34c26-92a0-4620-f53d-0277f6bc7d09", "colab": { "base_uri": "https://localhost:8080/", "height": 158 } }, "source": [ "data = np.zeros((4,10))\n", "i = 0\n", "for optimizer in models.keys():\n", " j = 0\n", " isTest = False\n", " for activation in activations:\n", " data[j, i] = models[optimizer][activation]['train'][-1]\n", " j += 1\n", " j = 0\n", " i += 1\n", " for activation in activations:\n", " data[j, i] = models[optimizer][activation]['test'][0]\n", " j += 1\n", " i += 1\n", "data" ], "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "array([[0.58575171, 0.55705822, 0.9155333 , 0.8134082 , 1.09731817,\n", " 1.0005126 , 0.31557649, 0.39956024, 0.311353 , 0.42637646],\n", " [0.54603666, 0.55165982, 0.69864362, 0.64264464, 1.26182771,\n", " 1.16288579, 0.25816193, 0.4195244 , 0.21717222, 0.37788409],\n", " [0.46336603, 0.49723259, 0.86144572, 0.77469265, 1.24011803,\n", " 1.12515199, 0.2693308 , 0.37503961, 0.24791728, 0.37985611],\n", " [0.36771712, 0.43835771, 0.81239867, 0.75216007, 1.27089512,\n", " 1.13645124, 0.17281628, 0.31153476, 0.14900836, 0.31503132]])" ] }, "metadata": { "tags": [] }, "execution_count": 68 } ] }, { "cell_type": "code", "metadata": { "id": "FcPEJO9eqjxX", "outputId": "a42808be-7865-45a7-9ab8-e12c0c580495", "colab": { "base_uri": "https://localhost:8080/", "height": 197 } }, "source": [ "models_df = pd.DataFrame(data=data, columns=index, index=activations)\n", "models_df" ], "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
OptimizerSGDAdaGradAdadeltaRMSpropAdam
Train/TestTrainingTestTrainingTestTrainingTestTrainingTestTrainingTest
sigmoid0.5857520.5570580.9155330.8134081.0973181.0005130.3155760.3995600.3113530.426376
tanh0.5460370.5516600.6986440.6426451.2618281.1628860.2581620.4195240.2171720.377884
softplus0.4633660.4972330.8614460.7746931.2401181.1251520.2693310.3750400.2479170.379856
relu0.3677170.4383580.8123990.7521601.2708951.1364510.1728160.3115350.1490080.315031
\n", "
" ], "text/plain": [ "Optimizer SGD AdaGrad Adadelta \\\n", "Train/Test Training Test Training Test Training Test \n", "sigmoid 0.585752 0.557058 0.915533 0.813408 1.097318 1.000513 \n", "tanh 0.546037 0.551660 0.698644 0.642645 1.261828 1.162886 \n", "softplus 0.463366 0.497233 0.861446 0.774693 1.240118 1.125152 \n", "relu 0.367717 0.438358 0.812399 0.752160 1.270895 1.136451 \n", "\n", "Optimizer RMSprop Adam \n", "Train/Test Training Test Training Test \n", "sigmoid 0.315576 0.399560 0.311353 0.426376 \n", "tanh 0.258162 0.419524 0.217172 0.377884 \n", "softplus 0.269331 0.375040 0.247917 0.379856 \n", "relu 0.172816 0.311535 0.149008 0.315031 " ] }, "metadata": { "tags": [] }, "execution_count": 69 } ] }, { "cell_type": "markdown", "metadata": { "id": "6SqAcQ9nqmzv" }, "source": [ "As can be seen, the best models were obtained with the Adam and the RMSprop optimizers, with ReLU as activation" ] }, { "cell_type": "markdown", "metadata": { "id": "sI2M1sFipzXt" }, "source": [ "# Your Homework \n", "\n", " **Other architectures - Multilayer Neural Networks** \n", "\n", "\n", " Up to now, the models had only one hidden layer. We will now experiment with more hidden layers, more specifically with 2 and 3 layers. Since the training for these kind of networks is very computationally expensive, only the case for models with the Adam optimizer and softplus as activation will be shown, as they rendered good results in previous models. Hence, the architectures will be (10-20-18), (10-20-20-18), (10-20-20-20-18).\n", "\n" ] } ] }