03_regression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.layers.experimental import preprocessing\n",
    "\n",
    "os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'\n",
    "np.set_printoptions(precision=3, suppress=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "url = 'http://archive.ics.uci.edu/ml/machine-learning-databases/auto-mpg/auto-mpg.data'\n",
    "column_names = ['MPG', 'Cylinders', 'Displacement', 'Horsepower', 'Weight',\n",
    "                'Acceleration', 'Model Year', 'Origin']\n",
    "\n",
    "dataset = pd.read_csv(url, names=column_names, na_values='?',\n",
    "                      comment='\\t', sep=' ', skipinitialspace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "outputs": [
    {
     "data": {
      "text/plain": "      MPG  Cylinders  Displacement  Horsepower  Weight  Acceleration  \\\n393  27.0          4         140.0        86.0  2790.0          15.6   \n394  44.0          4          97.0        52.0  2130.0          24.6   \n395  32.0          4         135.0        84.0  2295.0          11.6   \n396  28.0          4         120.0        79.0  2625.0          18.6   \n397  31.0          4         119.0        82.0  2720.0          19.4   \n\n     Model Year  Origin  \n393          82       1  \n394          82       2  \n395          82       1  \n396          82       1  \n397          82       1  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>MPG</th>\n      <th>Cylinders</th>\n      <th>Displacement</th>\n      <th>Horsepower</th>\n      <th>Weight</th>\n      <th>Acceleration</th>\n      <th>Model Year</th>\n      <th>Origin</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>393</th>\n      <td>27.0</td>\n      <td>4</td>\n      <td>140.0</td>\n      <td>86.0</td>\n      <td>2790.0</td>\n      <td>15.6</td>\n      <td>82</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>394</th>\n      <td>44.0</td>\n      <td>4</td>\n      <td>97.0</td>\n      <td>52.0</td>\n      <td>2130.0</td>\n      <td>24.6</td>\n      <td>82</td>\n      <td>2</td>\n    </tr>\n    <tr>\n      <th>395</th>\n      <td>32.0</td>\n      <td>4</td>\n      <td>135.0</td>\n      <td>84.0</td>\n      <td>2295.0</td>\n      <td>11.6</td>\n      <td>82</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>396</th>\n      <td>28.0</td>\n      <td>4</td>\n      <td>120.0</td>\n      <td>79.0</td>\n      <td>2625.0</td>\n      <td>18.6</td>\n      <td>82</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>397</th>\n      <td>31.0</td>\n      <td>4</td>\n      <td>119.0</td>\n      <td>82.0</td>\n      <td>2720.0</td>\n      <td>19.4</td>\n      <td>82</td>\n      <td>1</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.tail()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "outputs": [],
   "source": [
    "dataset = dataset.dropna()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "outputs": [
    {
     "data": {
      "text/plain": "      MPG  Cylinders  Displacement  Horsepower  Weight  Acceleration  \\\n393  27.0          4         140.0        86.0  2790.0          15.6   \n394  44.0          4          97.0        52.0  2130.0          24.6   \n395  32.0          4         135.0        84.0  2295.0          11.6   \n396  28.0          4         120.0        79.0  2625.0          18.6   \n397  31.0          4         119.0        82.0  2720.0          19.4   \n\n     Model Year  USA  Europe  Japan  \n393          82    1       0      0  \n394          82    0       1      0  \n395          82    1       0      0  \n396          82    1       0      0  \n397          82    1       0      0  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>MPG</th>\n      <th>Cylinders</th>\n      <th>Displacement</th>\n      <th>Horsepower</th>\n      <th>Weight</th>\n      <th>Acceleration</th>\n      <th>Model Year</th>\n      <th>USA</th>\n      <th>Europe</th>\n      <th>Japan</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>393</th>\n      <td>27.0</td>\n      <td>4</td>\n      <td>140.0</td>\n      <td>86.0</td>\n      <td>2790.0</td>\n      <td>15.6</td>\n      <td>82</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>394</th>\n      <td>44.0</td>\n      <td>4</td>\n      <td>97.0</td>\n      <td>52.0</td>\n      <td>2130.0</td>\n      <td>24.6</td>\n      <td>82</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>395</th>\n      <td>32.0</td>\n      <td>4</td>\n      <td>135.0</td>\n      <td>84.0</td>\n      <td>2295.0</td>\n      <td>11.6</td>\n      <td>82</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>396</th>\n      <td>28.0</td>\n      <td>4</td>\n      <td>120.0</td>\n      <td>79.0</td>\n      <td>2625.0</td>\n      <td>18.6</td>\n      <td>82</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>397</th>\n      <td>31.0</td>\n      <td>4</td>\n      <td>119.0</td>\n      <td>82.0</td>\n      <td>2720.0</td>\n      <td>19.4</td>\n      <td>82</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "origin = dataset.pop('Origin')\n",
    "dataset['USA'] = (origin == 1)*1\n",
    "dataset['Europe'] = (origin==2)*1\n",
    "dataset['Japan'] = (origin == 3)*1\n",
    "dataset.tail()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(392, 10) (314, 10) (78, 10)\n"
     ]
    }
   ],
   "source": [
    "train_dataset = dataset.sample(frac=0.8, random_state=0)\n",
    "test_dataset = dataset.drop(train_dataset.index)\n",
    "\n",
    "print(dataset.shape, train_dataset.shape, test_dataset.shape)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [
    {
     "data": {
      "text/plain": "              count         mean         std     min      25%     50%  \\\nMPG           314.0    23.310510    7.728652    10.0    17.00    22.0   \nCylinders     314.0     5.477707    1.699788     3.0     4.00     4.0   \nDisplacement  314.0   195.318471  104.331589    68.0   105.50   151.0   \nHorsepower    314.0   104.869427   38.096214    46.0    76.25    94.5   \nWeight        314.0  2990.251592  843.898596  1649.0  2256.50  2822.5   \nAcceleration  314.0    15.559236    2.789230     8.0    13.80    15.5   \nModel Year    314.0    75.898089    3.675642    70.0    73.00    76.0   \nUSA           314.0     0.624204    0.485101     0.0     0.00     1.0   \nEurope        314.0     0.178344    0.383413     0.0     0.00     0.0   \nJapan         314.0     0.197452    0.398712     0.0     0.00     0.0   \n\n                  75%     max  \nMPG             28.95    46.6  \nCylinders        8.00     8.0  \nDisplacement   265.75   455.0  \nHorsepower     128.00   225.0  \nWeight        3608.00  5140.0  \nAcceleration    17.20    24.8  \nModel Year      79.00    82.0  \nUSA              1.00     1.0  \nEurope           0.00     1.0  \nJapan            0.00     1.0  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>count</th>\n      <th>mean</th>\n      <th>std</th>\n      <th>min</th>\n      <th>25%</th>\n      <th>50%</th>\n      <th>75%</th>\n      <th>max</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>MPG</th>\n      <td>314.0</td>\n      <td>23.310510</td>\n      <td>7.728652</td>\n      <td>10.0</td>\n      <td>17.00</td>\n      <td>22.0</td>\n      <td>28.95</td>\n      <td>46.6</td>\n    </tr>\n    <tr>\n      <th>Cylinders</th>\n      <td>314.0</td>\n      <td>5.477707</td>\n      <td>1.699788</td>\n      <td>3.0</td>\n      <td>4.00</td>\n      <td>4.0</td>\n      <td>8.00</td>\n      <td>8.0</td>\n    </tr>\n    <tr>\n      <th>Displacement</th>\n      <td>314.0</td>\n      <td>195.318471</td>\n      <td>104.331589</td>\n      <td>68.0</td>\n      <td>105.50</td>\n      <td>151.0</td>\n      <td>265.75</td>\n      <td>455.0</td>\n    </tr>\n    <tr>\n      <th>Horsepower</th>\n      <td>314.0</td>\n      <td>104.869427</td>\n      <td>38.096214</td>\n      <td>46.0</td>\n      <td>76.25</td>\n      <td>94.5</td>\n      <td>128.00</td>\n      <td>225.0</td>\n    </tr>\n    <tr>\n      <th>Weight</th>\n      <td>314.0</td>\n      <td>2990.251592</td>\n      <td>843.898596</td>\n      <td>1649.0</td>\n      <td>2256.50</td>\n      <td>2822.5</td>\n      <td>3608.00</td>\n      <td>5140.0</td>\n    </tr>\n    <tr>\n      <th>Acceleration</th>\n      <td>314.0</td>\n      <td>15.559236</td>\n      <td>2.789230</td>\n      <td>8.0</td>\n      <td>13.80</td>\n      <td>15.5</td>\n      <td>17.20</td>\n      <td>24.8</td>\n    </tr>\n    <tr>\n      <th>Model Year</th>\n      <td>314.0</td>\n      <td>75.898089</td>\n      <td>3.675642</td>\n      <td>70.0</td>\n      <td>73.00</td>\n      <td>76.0</td>\n      <td>79.00</td>\n      <td>82.0</td>\n    </tr>\n    <tr>\n      <th>USA</th>\n      <td>314.0</td>\n      <td>0.624204</td>\n      <td>0.485101</td>\n      <td>0.0</td>\n      <td>0.00</td>\n      <td>1.0</td>\n      <td>1.00</td>\n      <td>1.0</td>\n    </tr>\n    <tr>\n      <th>Europe</th>\n      <td>314.0</td>\n      <td>0.178344</td>\n      <td>0.383413</td>\n      <td>0.0</td>\n      <td>0.00</td>\n      <td>0.0</td>\n      <td>0.00</td>\n      <td>1.0</td>\n    </tr>\n    <tr>\n      <th>Japan</th>\n      <td>314.0</td>\n      <td>0.197452</td>\n      <td>0.398712</td>\n      <td>0.0</td>\n      <td>0.00</td>\n      <td>0.0</td>\n      <td>0.00</td>\n      <td>1.0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dataset.describe().transpose()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "outputs": [],
   "source": [
    "train_x = train_dataset.copy()\n",
    "train_y = train_x.pop('MPG')\n",
    "\n",
    "test_x = test_dataset.copy()\n",
    "test_y = test_x.pop('MPG')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "outputs": [],
   "source": [
    "def plot(feature, x=None, y=None):\n",
    "    plt.figure(figsize=(10, 8))\n",
    "    plt.scatter(train_x[feature], train_y, label='Data')\n",
    "    if x is not None and y is not None:\n",
    "        plt.plot(x, y, color='k', label='Predictions')\n",
    "    plt.xlabel(feature)\n",
    "    plt.ylabel('MPG')\n",
    "    plt.legend()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 720x576 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot('Horsepower')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 720x576 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot('Weight')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 720x576 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot('Cylinders')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                     mean         std\n",
      "MPG             23.310510    7.728652\n",
      "Cylinders        5.477707    1.699788\n",
      "Displacement   195.318471  104.331589\n",
      "Horsepower     104.869427   38.096214\n",
      "Weight        2990.251592  843.898596\n",
      "Acceleration    15.559236    2.789230\n",
      "Model Year      75.898089    3.675642\n",
      "USA              0.624204    0.485101\n",
      "Europe           0.178344    0.383413\n",
      "Japan            0.197452    0.398712\n"
     ]
    }
   ],
   "source": [
    "print(train_dataset.describe().transpose()[['mean', 'std']])"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Normalization with Keras Preprocessing Normalization Class\n",
    "Normalization works like:\n",
    "$normalized input feature value = (input - mean / std)$"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[   5.478  195.318  104.869 2990.252   15.559   75.898    0.624    0.178\n",
      "     0.197]]\n"
     ]
    }
   ],
   "source": [
    "# Normalize\n",
    "normalizer = preprocessing.Normalization()\n",
    "normalizer.adapt(np.array(train_x))\n",
    "print(normalizer.mean.numpy())"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Raw example:  [[   4.    90.    75.  2125.    14.5   74.     1.     0.     0. ]]\n",
      "Normalized example:  [[-0.871 -1.011 -0.785 -1.027 -0.38  -0.517  0.776 -0.466 -0.496]]\n"
     ]
    }
   ],
   "source": [
    "first = np.array(train_x[:1])\n",
    "print('Raw example: ', first)\n",
    "print('Normalized example: ', normalizer(first).numpy())"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Regression\n",
    "$y = mx + b$"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(314,) (314, 9)\n"
     ]
    }
   ],
   "source": [
    "feature = 'Weight'\n",
    "single_feature = np.array(train_x[feature])\n",
    "print(single_feature.shape, train_x.shape)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "outputs": [],
   "source": [
    "single_feature_normalizer = preprocessing.Normalization(input_shape=[1,], axis=None)\n",
    "single_feature_normalizer.adapt(single_feature)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization_1 (Normalizat  (None, 1)                3         \n",
      " ion)                                                            \n",
      "                                                                 \n",
      " dense (Dense)               (None, 1)                 2         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 5\n",
      "Trainable params: 2\n",
      "Non-trainable params: 3\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "single_feature_model = keras.models.Sequential([\n",
    "    single_feature_normalizer,\n",
    "    layers.Dense(units=1) # Linear Model\n",
    "])\n",
    "\n",
    "single_feature_model.summary()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Loss Functions\n",
    "Mean Squared Error:\n",
    "Mean Absolute Error:"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "outputs": [],
   "source": [
    "loss = keras.losses.MeanAbsoluteError()\n",
    "optimizer = keras.optimizers.Adam(learning_rate=0.1)\n",
    "\n",
    "single_feature_model.compile(optimizer, loss)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "8/8 [==============================] - 0s 17ms/step - loss: 22.8903 - val_loss: 22.9151\n",
      "Epoch 2/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 22.0667 - val_loss: 22.1140\n",
      "Epoch 3/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 21.2670 - val_loss: 21.3141\n",
      "Epoch 4/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 20.4717 - val_loss: 20.5128\n",
      "Epoch 5/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 19.6726 - val_loss: 19.7155\n",
      "Epoch 6/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 18.8664 - val_loss: 18.9154\n",
      "Epoch 7/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 18.0667 - val_loss: 18.1149\n",
      "Epoch 8/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 17.2672 - val_loss: 17.3158\n",
      "Epoch 9/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 16.4668 - val_loss: 16.5166\n",
      "Epoch 10/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 15.6660 - val_loss: 15.7165\n",
      "Epoch 11/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 14.8700 - val_loss: 14.9151\n",
      "Epoch 12/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 14.0724 - val_loss: 14.1220\n",
      "Epoch 13/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 13.2953 - val_loss: 13.3403\n",
      "Epoch 14/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 12.5173 - val_loss: 12.5747\n",
      "Epoch 15/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 11.7435 - val_loss: 11.8160\n",
      "Epoch 16/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 10.9655 - val_loss: 11.0606\n",
      "Epoch 17/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 10.1850 - val_loss: 10.3099\n",
      "Epoch 18/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 9.4292 - val_loss: 9.5633\n",
      "Epoch 19/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 8.6670 - val_loss: 8.8457\n",
      "Epoch 20/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 7.9231 - val_loss: 8.1376\n",
      "Epoch 21/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 7.1965 - val_loss: 7.4402\n",
      "Epoch 22/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 6.4970 - val_loss: 6.7552\n",
      "Epoch 23/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 5.8392 - val_loss: 6.0990\n",
      "Epoch 24/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 5.2529 - val_loss: 5.4989\n",
      "Epoch 25/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 4.7163 - val_loss: 4.9686\n",
      "Epoch 26/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 4.2935 - val_loss: 4.5407\n",
      "Epoch 27/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.9384 - val_loss: 4.2276\n",
      "Epoch 28/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.6723 - val_loss: 3.9807\n",
      "Epoch 29/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.5101 - val_loss: 3.7958\n",
      "Epoch 30/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.4233 - val_loss: 3.6652\n",
      "Epoch 31/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.3665 - val_loss: 3.5663\n",
      "Epoch 32/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.3315 - val_loss: 3.5137\n",
      "Epoch 33/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.3159 - val_loss: 3.4697\n",
      "Epoch 34/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2992 - val_loss: 3.4444\n",
      "Epoch 35/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2912 - val_loss: 3.4283\n",
      "Epoch 36/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2833 - val_loss: 3.4228\n",
      "Epoch 37/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2836 - val_loss: 3.4174\n",
      "Epoch 38/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2846 - val_loss: 3.4124\n",
      "Epoch 39/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2801 - val_loss: 3.4069\n",
      "Epoch 40/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2824 - val_loss: 3.4058\n",
      "Epoch 41/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2785 - val_loss: 3.4046\n",
      "Epoch 42/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2785 - val_loss: 3.4031\n",
      "Epoch 43/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2831 - val_loss: 3.4052\n",
      "Epoch 44/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2765 - val_loss: 3.4041\n",
      "Epoch 45/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2793 - val_loss: 3.4003\n",
      "Epoch 46/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2783 - val_loss: 3.3984\n",
      "Epoch 47/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2789 - val_loss: 3.3976\n",
      "Epoch 48/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2804 - val_loss: 3.3960\n",
      "Epoch 49/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2785 - val_loss: 3.3946\n",
      "Epoch 50/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2802 - val_loss: 3.3937\n",
      "Epoch 51/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2815 - val_loss: 3.3996\n",
      "Epoch 52/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2802 - val_loss: 3.3956\n",
      "Epoch 53/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2779 - val_loss: 3.3931\n",
      "Epoch 54/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2775 - val_loss: 3.3914\n",
      "Epoch 55/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2770 - val_loss: 3.3903\n",
      "Epoch 56/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2782 - val_loss: 3.3912\n",
      "Epoch 57/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2797 - val_loss: 3.3944\n",
      "Epoch 58/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2796 - val_loss: 3.3909\n",
      "Epoch 59/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2816 - val_loss: 3.3885\n",
      "Epoch 60/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2789 - val_loss: 3.3863\n",
      "Epoch 61/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2785 - val_loss: 3.3865\n",
      "Epoch 62/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2782 - val_loss: 3.3865\n",
      "Epoch 63/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2792 - val_loss: 3.3882\n",
      "Epoch 64/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2776 - val_loss: 3.3872\n",
      "Epoch 65/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2766 - val_loss: 3.3860\n",
      "Epoch 66/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2789 - val_loss: 3.3849\n",
      "Epoch 67/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2783 - val_loss: 3.3878\n",
      "Epoch 68/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2797 - val_loss: 3.3913\n",
      "Epoch 69/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2787 - val_loss: 3.3881\n",
      "Epoch 70/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2768 - val_loss: 3.3886\n",
      "Epoch 71/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2775 - val_loss: 3.3871\n",
      "Epoch 72/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2797 - val_loss: 3.3847\n",
      "Epoch 73/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2776 - val_loss: 3.3857\n",
      "Epoch 74/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2777 - val_loss: 3.3850\n",
      "Epoch 75/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2827 - val_loss: 3.3906\n",
      "Epoch 76/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2805 - val_loss: 3.3852\n",
      "Epoch 77/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2852 - val_loss: 3.3853\n",
      "Epoch 78/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2853 - val_loss: 3.3804\n",
      "Epoch 79/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2776 - val_loss: 3.3818\n",
      "Epoch 80/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2822 - val_loss: 3.3908\n",
      "Epoch 81/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2777 - val_loss: 3.3879\n",
      "Epoch 82/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2765 - val_loss: 3.3862\n",
      "Epoch 83/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2871 - val_loss: 3.3869\n",
      "Epoch 84/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2814 - val_loss: 3.3815\n",
      "Epoch 85/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2877 - val_loss: 3.3853\n",
      "Epoch 86/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2780 - val_loss: 3.3871\n",
      "Epoch 87/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2758 - val_loss: 3.3874\n",
      "Epoch 88/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2770 - val_loss: 3.3871\n",
      "Epoch 89/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2791 - val_loss: 3.3865\n",
      "Epoch 90/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2779 - val_loss: 3.3857\n",
      "Epoch 91/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2763 - val_loss: 3.3880\n",
      "Epoch 92/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2779 - val_loss: 3.3846\n",
      "Epoch 93/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2836 - val_loss: 3.3877\n",
      "Epoch 94/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2787 - val_loss: 3.3840\n",
      "Epoch 95/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.2774 - val_loss: 3.3870\n",
      "Epoch 96/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2805 - val_loss: 3.3840\n",
      "Epoch 97/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2768 - val_loss: 3.3865\n",
      "Epoch 98/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2805 - val_loss: 3.3889\n",
      "Epoch 99/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2790 - val_loss: 3.3883\n",
      "Epoch 100/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2780 - val_loss: 3.3870\n"
     ]
    }
   ],
   "source": [
    "history = single_feature_model.fit(train_x[feature], train_y, epochs=100, verbose=1, validation_split=0.2)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_loss(history):\n",
    "    plt.plot(history.history['loss'], label='loss')\n",
    "    plt.plot(history.history['val_loss'], label='validation loss')\n",
    "    plt.ylim([0, 50])\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Error | MPG')\n",
    "    plt.legend()\n",
    "    plt.grid(True)\n",
    "\n",
    "plot_loss(history)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3/3 [==============================] - 0s 2ms/step - loss: 3.1082\n"
     ]
    },
    {
     "data": {
      "text/plain": "3.1082496643066406"
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "single_feature_model.evaluate(test_x[feature], test_y, verbose=1)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7/7 [==============================] - 0s 1ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 720x576 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "range_min = np.min(test_x[feature]) - 10\n",
    "range_max = np.max(test_x[feature]) + 10\n",
    "x = tf.linspace(range_min, range_max, 200)\n",
    "y = single_feature_model.predict(x)\n",
    "\n",
    "plot(feature, x, y)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization_1 (Normalizat  (None, 1)                3         \n",
      " ion)                                                            \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 64)                128       \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 64)                4160      \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 1)                 65        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 4,356\n",
      "Trainable params: 4,353\n",
      "Non-trainable params: 3\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "dnn_model = keras.Sequential([\n",
    "    single_feature_normalizer,\n",
    "    layers.Dense(64, activation='relu'),\n",
    "    layers.Dense(64, activation='relu'),\n",
    "    layers.Dense(1)\n",
    "])\n",
    "\n",
    "loss = keras.losses.MeanAbsoluteError()\n",
    "optimizer = keras.optimizers.Adam(learning_rate=0.001)\n",
    "\n",
    "dnn_model.compile(optimizer, loss)\n",
    "\n",
    "dnn_model.summary()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "8/8 [==============================] - 0s 16ms/step - loss: 23.0910 - val_loss: 23.4585\n",
      "Epoch 2/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 22.8226 - val_loss: 23.1651\n",
      "Epoch 3/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 22.5157 - val_loss: 22.8272\n",
      "Epoch 4/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 22.1489 - val_loss: 22.3941\n",
      "Epoch 5/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 21.6749 - val_loss: 21.8331\n",
      "Epoch 6/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 21.0684 - val_loss: 21.1050\n",
      "Epoch 7/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 20.2738 - val_loss: 20.1696\n",
      "Epoch 8/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 19.2477 - val_loss: 18.9970\n",
      "Epoch 9/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 17.9838 - val_loss: 17.5362\n",
      "Epoch 10/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 16.4020 - val_loss: 15.8433\n",
      "Epoch 11/100\n",
      "8/8 [==============================] - 0s 6ms/step - loss: 14.5957 - val_loss: 14.0085\n",
      "Epoch 12/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 12.5683 - val_loss: 12.0513\n",
      "Epoch 13/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 10.5188 - val_loss: 9.9827\n",
      "Epoch 14/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 8.4054 - val_loss: 7.8447\n",
      "Epoch 15/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 6.4270 - val_loss: 6.1026\n",
      "Epoch 16/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 5.3406 - val_loss: 5.3066\n",
      "Epoch 17/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 5.0734 - val_loss: 5.0436\n",
      "Epoch 18/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 4.9171 - val_loss: 4.8546\n",
      "Epoch 19/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 4.7210 - val_loss: 4.7007\n",
      "Epoch 20/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 4.4918 - val_loss: 4.5781\n",
      "Epoch 21/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 4.3208 - val_loss: 4.5009\n",
      "Epoch 22/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 4.1645 - val_loss: 4.3621\n",
      "Epoch 23/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 4.0169 - val_loss: 4.1739\n",
      "Epoch 24/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.8959 - val_loss: 4.0112\n",
      "Epoch 25/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.8017 - val_loss: 3.8979\n",
      "Epoch 26/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.7138 - val_loss: 3.8088\n",
      "Epoch 27/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.6412 - val_loss: 3.7327\n",
      "Epoch 28/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.5833 - val_loss: 3.6846\n",
      "Epoch 29/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.5293 - val_loss: 3.6167\n",
      "Epoch 30/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.4926 - val_loss: 3.5357\n",
      "Epoch 31/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.4423 - val_loss: 3.4849\n",
      "Epoch 32/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.4000 - val_loss: 3.4493\n",
      "Epoch 33/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.3768 - val_loss: 3.3964\n",
      "Epoch 34/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.3471 - val_loss: 3.3706\n",
      "Epoch 35/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.3189 - val_loss: 3.3482\n",
      "Epoch 36/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.3040 - val_loss: 3.3350\n",
      "Epoch 37/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2765 - val_loss: 3.3097\n",
      "Epoch 38/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.2574 - val_loss: 3.2867\n",
      "Epoch 39/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2409 - val_loss: 3.2775\n",
      "Epoch 40/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2302 - val_loss: 3.2633\n",
      "Epoch 41/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2193 - val_loss: 3.2594\n",
      "Epoch 42/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.2050 - val_loss: 3.2349\n",
      "Epoch 43/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.1939 - val_loss: 3.2311\n",
      "Epoch 44/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1839 - val_loss: 3.2221\n",
      "Epoch 45/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1755 - val_loss: 3.2160\n",
      "Epoch 46/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1711 - val_loss: 3.2033\n",
      "Epoch 47/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1652 - val_loss: 3.1936\n",
      "Epoch 48/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1581 - val_loss: 3.1989\n",
      "Epoch 49/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1560 - val_loss: 3.1781\n",
      "Epoch 50/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1497 - val_loss: 3.1717\n",
      "Epoch 51/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.1539 - val_loss: 3.1848\n",
      "Epoch 52/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1488 - val_loss: 3.1599\n",
      "Epoch 53/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1353 - val_loss: 3.1555\n",
      "Epoch 54/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1357 - val_loss: 3.1537\n",
      "Epoch 55/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1337 - val_loss: 3.1402\n",
      "Epoch 56/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1356 - val_loss: 3.1379\n",
      "Epoch 57/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1282 - val_loss: 3.1325\n",
      "Epoch 58/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1383 - val_loss: 3.1516\n",
      "Epoch 59/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1367 - val_loss: 3.1219\n",
      "Epoch 60/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1251 - val_loss: 3.1259\n",
      "Epoch 61/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.1321 - val_loss: 3.1251\n",
      "Epoch 62/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1381 - val_loss: 3.1131\n",
      "Epoch 63/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1224 - val_loss: 3.1197\n",
      "Epoch 64/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1221 - val_loss: 3.1245\n",
      "Epoch 65/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1150 - val_loss: 3.1080\n",
      "Epoch 66/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1273 - val_loss: 3.1024\n",
      "Epoch 67/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1240 - val_loss: 3.1093\n",
      "Epoch 68/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1207 - val_loss: 3.1394\n",
      "Epoch 69/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1227 - val_loss: 3.1019\n",
      "Epoch 70/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1149 - val_loss: 3.0989\n",
      "Epoch 71/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1202 - val_loss: 3.0973\n",
      "Epoch 72/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 3.1104 - val_loss: 3.1066\n",
      "Epoch 73/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1119 - val_loss: 3.0955\n",
      "Epoch 74/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1123 - val_loss: 3.0990\n",
      "Epoch 75/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1128 - val_loss: 3.0954\n",
      "Epoch 76/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1064 - val_loss: 3.0906\n",
      "Epoch 77/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1262 - val_loss: 3.0999\n",
      "Epoch 78/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1098 - val_loss: 3.0819\n",
      "Epoch 79/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1054 - val_loss: 3.0863\n",
      "Epoch 80/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1060 - val_loss: 3.1023\n",
      "Epoch 81/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1058 - val_loss: 3.0817\n",
      "Epoch 82/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1038 - val_loss: 3.0790\n",
      "Epoch 83/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1181 - val_loss: 3.1245\n",
      "Epoch 84/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1192 - val_loss: 3.0706\n",
      "Epoch 85/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1006 - val_loss: 3.0780\n",
      "Epoch 86/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1046 - val_loss: 3.0872\n",
      "Epoch 87/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0952 - val_loss: 3.0713\n",
      "Epoch 88/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.0976 - val_loss: 3.0682\n",
      "Epoch 89/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 3.1011 - val_loss: 3.0754\n",
      "Epoch 90/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1065 - val_loss: 3.0872\n",
      "Epoch 91/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0907 - val_loss: 3.0634\n",
      "Epoch 92/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0971 - val_loss: 3.0616\n",
      "Epoch 93/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0917 - val_loss: 3.0858\n",
      "Epoch 94/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0955 - val_loss: 3.0656\n",
      "Epoch 95/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1006 - val_loss: 3.0589\n",
      "Epoch 96/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0955 - val_loss: 3.0657\n",
      "Epoch 97/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1022 - val_loss: 3.0897\n",
      "Epoch 98/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0917 - val_loss: 3.0602\n",
      "Epoch 99/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0897 - val_loss: 3.0634\n",
      "Epoch 100/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0869 - val_loss: 3.0662\n"
     ]
    }
   ],
   "source": [
    "history = dnn_model.fit(train_x[feature], train_y, validation_split=0.2, verbose=1, epochs=100)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3/3 [==============================] - 0s 2ms/step - loss: 2.7518\n"
     ]
    },
    {
     "data": {
      "text/plain": "2.751774787902832"
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dnn_model.evaluate(test_x[feature], test_y, verbose=1)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7/7 [==============================] - 0s 1ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 720x576 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = tf.linspace(range_min, range_max, 200)\n",
    "y = single_feature_model.predict(x)\n",
    "\n",
    "plot(feature, x, y)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization (Normalizatio  (None, 9)                19        \n",
      " n)                                                              \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 1)                 10        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 29\n",
      "Trainable params: 10\n",
      "Non-trainable params: 19\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "linear_model = keras.Sequential([\n",
    "    normalizer,\n",
    "    layers.Dense(1)\n",
    "])\n",
    "\n",
    "loss = keras.losses.MeanAbsoluteError()\n",
    "optimizer = keras.optimizers.Adam(learning_rate=0.1)\n",
    "\n",
    "linear_model.compile(optimizer, loss)\n",
    "\n",
    "linear_model.summary()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "8/8 [==============================] - 0s 16ms/step - loss: 22.9405 - val_loss: 22.9865\n",
      "Epoch 2/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 22.0587 - val_loss: 22.2663\n",
      "Epoch 3/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 21.2476 - val_loss: 21.5680\n",
      "Epoch 4/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 20.4104 - val_loss: 20.8296\n",
      "Epoch 5/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 19.5951 - val_loss: 20.1234\n",
      "Epoch 6/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 18.7660 - val_loss: 19.3886\n",
      "Epoch 7/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 17.9454 - val_loss: 18.6585\n",
      "Epoch 8/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 17.1352 - val_loss: 17.9379\n",
      "Epoch 9/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 16.3358 - val_loss: 17.2309\n",
      "Epoch 10/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 15.5156 - val_loss: 16.4915\n",
      "Epoch 11/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 14.7166 - val_loss: 15.7883\n",
      "Epoch 12/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 13.9006 - val_loss: 15.0539\n",
      "Epoch 13/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 13.0895 - val_loss: 14.3281\n",
      "Epoch 14/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 12.3090 - val_loss: 13.5596\n",
      "Epoch 15/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 11.4848 - val_loss: 12.8613\n",
      "Epoch 16/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 10.7021 - val_loss: 12.1623\n",
      "Epoch 17/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 9.9014 - val_loss: 11.4538\n",
      "Epoch 18/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 9.1267 - val_loss: 10.6223\n",
      "Epoch 19/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 8.3715 - val_loss: 9.9132\n",
      "Epoch 20/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 7.6327 - val_loss: 9.2141\n",
      "Epoch 21/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 6.9805 - val_loss: 8.4670\n",
      "Epoch 22/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 6.3405 - val_loss: 7.8481\n",
      "Epoch 23/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 5.7944 - val_loss: 7.1822\n",
      "Epoch 24/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 5.2619 - val_loss: 6.5959\n",
      "Epoch 25/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 4.8304 - val_loss: 6.0245\n",
      "Epoch 26/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 4.3667 - val_loss: 5.3019\n",
      "Epoch 27/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.9108 - val_loss: 4.6891\n",
      "Epoch 28/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.5151 - val_loss: 3.9522\n",
      "Epoch 29/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.1191 - val_loss: 3.5318\n",
      "Epoch 30/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.9027 - val_loss: 3.1240\n",
      "Epoch 31/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.7864 - val_loss: 2.8765\n",
      "Epoch 32/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.6796 - val_loss: 2.7665\n",
      "Epoch 33/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5945 - val_loss: 2.7428\n",
      "Epoch 34/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5713 - val_loss: 2.6722\n",
      "Epoch 35/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5631 - val_loss: 2.6657\n",
      "Epoch 36/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5178 - val_loss: 2.6020\n",
      "Epoch 37/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5145 - val_loss: 2.6044\n",
      "Epoch 38/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4991 - val_loss: 2.5526\n",
      "Epoch 39/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5051 - val_loss: 2.5522\n",
      "Epoch 40/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4851 - val_loss: 2.5170\n",
      "Epoch 41/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5082 - val_loss: 2.5343\n",
      "Epoch 42/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4984 - val_loss: 2.5551\n",
      "Epoch 43/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.4950 - val_loss: 2.5090\n",
      "Epoch 44/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5036 - val_loss: 2.4964\n",
      "Epoch 45/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5265 - val_loss: 2.4906\n",
      "Epoch 46/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5031 - val_loss: 2.4957\n",
      "Epoch 47/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4992 - val_loss: 2.4979\n",
      "Epoch 48/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5673 - val_loss: 2.4904\n",
      "Epoch 49/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5000 - val_loss: 2.4821\n",
      "Epoch 50/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4680 - val_loss: 2.5756\n",
      "Epoch 51/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5033 - val_loss: 2.4811\n",
      "Epoch 52/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5257 - val_loss: 2.5476\n",
      "Epoch 53/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5021 - val_loss: 2.5111\n",
      "Epoch 54/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4973 - val_loss: 2.5122\n",
      "Epoch 55/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5158 - val_loss: 2.4703\n",
      "Epoch 56/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4689 - val_loss: 2.5353\n",
      "Epoch 57/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4723 - val_loss: 2.4549\n",
      "Epoch 58/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5003 - val_loss: 2.4996\n",
      "Epoch 59/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4811 - val_loss: 2.4601\n",
      "Epoch 60/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5339 - val_loss: 2.4798\n",
      "Epoch 61/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4809 - val_loss: 2.5229\n",
      "Epoch 62/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4768 - val_loss: 2.4910\n",
      "Epoch 63/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.4808 - val_loss: 2.5079\n",
      "Epoch 64/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4697 - val_loss: 2.4783\n",
      "Epoch 65/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4742 - val_loss: 2.5376\n",
      "Epoch 66/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4919 - val_loss: 2.4668\n",
      "Epoch 67/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5248 - val_loss: 2.5145\n",
      "Epoch 68/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.5327 - val_loss: 2.4684\n",
      "Epoch 69/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5032 - val_loss: 2.5394\n",
      "Epoch 70/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4969 - val_loss: 2.4650\n",
      "Epoch 71/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4793 - val_loss: 2.4975\n",
      "Epoch 72/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.4925 - val_loss: 2.4783\n",
      "Epoch 73/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4958 - val_loss: 2.4800\n",
      "Epoch 74/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4755 - val_loss: 2.4918\n",
      "Epoch 75/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4840 - val_loss: 2.4859\n",
      "Epoch 76/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4646 - val_loss: 2.4701\n",
      "Epoch 77/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5210 - val_loss: 2.5149\n",
      "Epoch 78/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5031 - val_loss: 2.4618\n",
      "Epoch 79/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4931 - val_loss: 2.5191\n",
      "Epoch 80/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5002 - val_loss: 2.4726\n",
      "Epoch 81/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 2.4695 - val_loss: 2.5314\n",
      "Epoch 82/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 2.4929 - val_loss: 2.4561\n",
      "Epoch 83/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 2.5130 - val_loss: 2.4700\n",
      "Epoch 84/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4757 - val_loss: 2.5080\n",
      "Epoch 85/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 2.4862 - val_loss: 2.4990\n",
      "Epoch 86/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4699 - val_loss: 2.5288\n",
      "Epoch 87/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4658 - val_loss: 2.4800\n",
      "Epoch 88/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4996 - val_loss: 2.4716\n",
      "Epoch 89/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5174 - val_loss: 2.4838\n",
      "Epoch 90/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4890 - val_loss: 2.4515\n",
      "Epoch 91/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.4840 - val_loss: 2.5298\n",
      "Epoch 92/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4994 - val_loss: 2.4965\n",
      "Epoch 93/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4712 - val_loss: 2.5167\n",
      "Epoch 94/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4849 - val_loss: 2.4634\n",
      "Epoch 95/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4822 - val_loss: 2.5320\n",
      "Epoch 96/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5137 - val_loss: 2.4463\n",
      "Epoch 97/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 2.5012 - val_loss: 2.5147\n",
      "Epoch 98/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5497 - val_loss: 2.4602\n",
      "Epoch 99/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4899 - val_loss: 2.5291\n",
      "Epoch 100/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.5399 - val_loss: 2.4555\n"
     ]
    }
   ],
   "source": [
    "history = linear_model.fit(train_x, train_y, epochs=100, verbose=1, validation_split=0.2)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3/3 [==============================] - 0s 2ms/step - loss: 2.4546\n"
     ]
    },
    {
     "data": {
      "text/plain": "2.454596519470215"
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "linear_model.evaluate(test_x, test_y, verbose=1)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_3\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " normalization (Normalizatio  (None, 9)                19        \n",
      " n)                                                              \n",
      "                                                                 \n",
      " dense_5 (Dense)             (None, 64)                640       \n",
      "                                                                 \n",
      " dense_6 (Dense)             (None, 64)                4160      \n",
      "                                                                 \n",
      " dense_7 (Dense)             (None, 1)                 65        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 4,884\n",
      "Trainable params: 4,865\n",
      "Non-trainable params: 19\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "dnn_model = keras.Sequential([\n",
    "    normalizer,\n",
    "    layers.Dense(64, activation='relu'),\n",
    "    layers.Dense(64, activation='relu'),\n",
    "    layers.Dense(1)\n",
    "])\n",
    "\n",
    "loss = keras.losses.MeanAbsoluteError()\n",
    "optimizer = keras.optimizers.Adam(learning_rate=0.001)\n",
    "\n",
    "dnn_model.compile(optimizer, loss)\n",
    "\n",
    "dnn_model.summary()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "8/8 [==============================] - 0s 16ms/step - loss: 23.0371 - val_loss: 23.1983\n",
      "Epoch 2/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 22.4519 - val_loss: 22.5227\n",
      "Epoch 3/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 21.7624 - val_loss: 21.6378\n",
      "Epoch 4/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 20.8269 - val_loss: 20.4598\n",
      "Epoch 5/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 19.5686 - val_loss: 18.8968\n",
      "Epoch 6/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 17.9474 - val_loss: 16.8700\n",
      "Epoch 7/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 15.9172 - val_loss: 14.7782\n",
      "Epoch 8/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 13.8961 - val_loss: 12.8501\n",
      "Epoch 9/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 11.7510 - val_loss: 10.5292\n",
      "Epoch 10/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 9.3512 - val_loss: 7.6858\n",
      "Epoch 11/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 6.8741 - val_loss: 5.8049\n",
      "Epoch 12/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 6.1115 - val_loss: 5.3953\n",
      "Epoch 13/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 5.4289 - val_loss: 4.7566\n",
      "Epoch 14/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 4.5807 - val_loss: 4.2801\n",
      "Epoch 15/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.9521 - val_loss: 3.9023\n",
      "Epoch 16/100\n",
      "8/8 [==============================] - 0s 6ms/step - loss: 3.5676 - val_loss: 3.6476\n",
      "Epoch 17/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.3240 - val_loss: 3.4615\n",
      "Epoch 18/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 3.0921 - val_loss: 3.2949\n",
      "Epoch 19/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.9256 - val_loss: 3.1120\n",
      "Epoch 20/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.8354 - val_loss: 2.9707\n",
      "Epoch 21/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.7095 - val_loss: 2.8778\n",
      "Epoch 22/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.6044 - val_loss: 2.8379\n",
      "Epoch 23/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4930 - val_loss: 2.6588\n",
      "Epoch 24/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.4241 - val_loss: 2.6152\n",
      "Epoch 25/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.3510 - val_loss: 2.5740\n",
      "Epoch 26/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.2931 - val_loss: 2.5155\n",
      "Epoch 27/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.2523 - val_loss: 2.4879\n",
      "Epoch 28/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.2046 - val_loss: 2.3723\n",
      "Epoch 29/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.1597 - val_loss: 2.3845\n",
      "Epoch 30/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.1054 - val_loss: 2.3951\n",
      "Epoch 31/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 2.0795 - val_loss: 2.3531\n",
      "Epoch 32/100\n",
      "8/8 [==============================] - 0s 6ms/step - loss: 2.0535 - val_loss: 2.3059\n",
      "Epoch 33/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 2.0277 - val_loss: 2.3114\n",
      "Epoch 34/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 2.0045 - val_loss: 2.2752\n",
      "Epoch 35/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 1.9816 - val_loss: 2.2567\n",
      "Epoch 36/100\n",
      "8/8 [==============================] - 0s 6ms/step - loss: 1.9789 - val_loss: 2.2611\n",
      "Epoch 37/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.9715 - val_loss: 2.1753\n",
      "Epoch 38/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.9573 - val_loss: 2.2473\n",
      "Epoch 39/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.9504 - val_loss: 2.2730\n",
      "Epoch 40/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.9143 - val_loss: 2.1955\n",
      "Epoch 41/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.9141 - val_loss: 2.2466\n",
      "Epoch 42/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.9125 - val_loss: 2.2662\n",
      "Epoch 43/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.9132 - val_loss: 2.1796\n",
      "Epoch 44/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8820 - val_loss: 2.2537\n",
      "Epoch 45/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8773 - val_loss: 2.1714\n",
      "Epoch 46/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8629 - val_loss: 2.2317\n",
      "Epoch 47/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8702 - val_loss: 2.1953\n",
      "Epoch 48/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 1.8817 - val_loss: 2.1753\n",
      "Epoch 49/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8572 - val_loss: 2.2505\n",
      "Epoch 50/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8676 - val_loss: 2.1720\n",
      "Epoch 51/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8477 - val_loss: 2.2362\n",
      "Epoch 52/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8413 - val_loss: 2.1784\n",
      "Epoch 53/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8561 - val_loss: 2.2208\n",
      "Epoch 54/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8161 - val_loss: 2.1450\n",
      "Epoch 55/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8130 - val_loss: 2.2622\n",
      "Epoch 56/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8086 - val_loss: 2.1643\n",
      "Epoch 57/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7987 - val_loss: 2.1775\n",
      "Epoch 58/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.8002 - val_loss: 2.2079\n",
      "Epoch 59/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 1.7876 - val_loss: 2.1639\n",
      "Epoch 60/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7709 - val_loss: 2.2694\n",
      "Epoch 61/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7901 - val_loss: 2.1756\n",
      "Epoch 62/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 1.7685 - val_loss: 2.2090\n",
      "Epoch 63/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7660 - val_loss: 2.1741\n",
      "Epoch 64/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 1.7616 - val_loss: 2.2166\n",
      "Epoch 65/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 1.7607 - val_loss: 2.1603\n",
      "Epoch 66/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7694 - val_loss: 2.2277\n",
      "Epoch 67/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7529 - val_loss: 2.1261\n",
      "Epoch 68/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7405 - val_loss: 2.2884\n",
      "Epoch 69/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7695 - val_loss: 2.1210\n",
      "Epoch 70/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7406 - val_loss: 2.2080\n",
      "Epoch 71/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7340 - val_loss: 2.2074\n",
      "Epoch 72/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7301 - val_loss: 2.1322\n",
      "Epoch 73/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7431 - val_loss: 2.2132\n",
      "Epoch 74/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7161 - val_loss: 2.1456\n",
      "Epoch 75/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7244 - val_loss: 2.2324\n",
      "Epoch 76/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7210 - val_loss: 2.1071\n",
      "Epoch 77/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 1.7248 - val_loss: 2.2119\n",
      "Epoch 78/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7018 - val_loss: 2.1554\n",
      "Epoch 79/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6993 - val_loss: 2.1632\n",
      "Epoch 80/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7104 - val_loss: 2.2130\n",
      "Epoch 81/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.7349 - val_loss: 2.1205\n",
      "Epoch 82/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6875 - val_loss: 2.2832\n",
      "Epoch 83/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 1.6974 - val_loss: 2.1444\n",
      "Epoch 84/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6867 - val_loss: 2.2676\n",
      "Epoch 85/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6849 - val_loss: 2.1452\n",
      "Epoch 86/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6739 - val_loss: 2.1641\n",
      "Epoch 87/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6767 - val_loss: 2.1664\n",
      "Epoch 88/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6697 - val_loss: 2.1732\n",
      "Epoch 89/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6668 - val_loss: 2.1664\n",
      "Epoch 90/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6729 - val_loss: 2.1814\n",
      "Epoch 91/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6739 - val_loss: 2.1714\n",
      "Epoch 92/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6538 - val_loss: 2.1244\n",
      "Epoch 93/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6626 - val_loss: 2.1821\n",
      "Epoch 94/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6628 - val_loss: 2.1884\n",
      "Epoch 95/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6478 - val_loss: 2.1383\n",
      "Epoch 96/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6485 - val_loss: 2.1545\n",
      "Epoch 97/100\n",
      "8/8 [==============================] - 0s 3ms/step - loss: 1.6396 - val_loss: 2.1401\n",
      "Epoch 98/100\n",
      "8/8 [==============================] - 0s 5ms/step - loss: 1.6391 - val_loss: 2.2273\n",
      "Epoch 99/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6411 - val_loss: 2.1525\n",
      "Epoch 100/100\n",
      "8/8 [==============================] - 0s 4ms/step - loss: 1.6485 - val_loss: 2.1525\n"
     ]
    }
   ],
   "source": [
    "history = dnn_model.fit(train_x, train_y, validation_split=0.2, verbose=1, epochs=100)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3/3 [==============================] - 0s 2ms/step - loss: 1.6481\n"
     ]
    },
    {
     "data": {
      "text/plain": "1.6480777263641357"
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dnn_model.evaluate(test_x, test_y, verbose=1)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.6 ('venv': venv)",
   "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": "3.9.6"
  },
  "vscode": {
   "interpreter": {
    "hash": "715d9a84b2ee7143435454b8a4c19aa708e808d01acd105a49ec5fb1f6b82df9"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}