1 - Simple linear regression in Keras.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1 - Simple regression model in TensorFlow 2.0\n",
    "\n",
    "In this notebook, we'll take a look at using TensorFlow 2.0 and the `tf.keras` interface to execute on a simple regression problem. We'll also take a look at the repeatability of machine learning training techniques and see if any uncertainty exists, and this example model isn't telling us about it."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Much of the material in this notebook is adapted from the TensorFlow 2.0 intro tutorial, available here: https://www.tensorflow.org/tutorials/keras/regression."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's get some data from the cars data set!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.0.0\n"
     ]
    }
   ],
   "source": [
    "import pathlib\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "\n",
    "tf.get_logger().setLevel('ERROR')\n",
    "\n",
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/Users/zach_anglin/.keras/datasets/auto-mpg.data'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset_path = tf.keras.utils.get_file(\"auto-mpg.data\", \"http://archive.ics.uci.edu/ml/machine-learning-databases/auto-mpg/auto-mpg.data\")\n",
    "dataset_path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "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",
       "      <td>393</td>\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",
       "      <td>394</td>\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",
       "      <td>395</td>\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",
       "      <td>396</td>\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",
       "      <td>397</td>\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>"
      ],
      "text/plain": [
       "      MPG  Cylinders  Displacement  Horsepower  Weight  Acceleration  \\\n",
       "393  27.0          4         140.0        86.0  2790.0          15.6   \n",
       "394  44.0          4          97.0        52.0  2130.0          24.6   \n",
       "395  32.0          4         135.0        84.0  2295.0          11.6   \n",
       "396  28.0          4         120.0        79.0  2625.0          18.6   \n",
       "397  31.0          4         119.0        82.0  2720.0          19.4   \n",
       "\n",
       "     Model Year  Origin  \n",
       "393          82       1  \n",
       "394          82       2  \n",
       "395          82       1  \n",
       "396          82       1  \n",
       "397          82       1  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "column_names = ['MPG','Cylinders','Displacement','Horsepower','Weight',\n",
    "                'Acceleration', 'Model Year', 'Origin']\n",
    "raw_dataset = pd.read_csv(dataset_path, names=column_names,\n",
    "                      na_values = \"?\", comment='\\t',\n",
    "                      sep=\" \", skipinitialspace=True)\n",
    "\n",
    "dataset = raw_dataset.copy()\n",
    "dataset.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Now that we have some data, we need to go about cleaning it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop NA values\n",
    "dataset = dataset.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We need to turn origin into a one-hot encoded vector.\n",
    "\n",
    "origin = dataset.pop('Origin')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "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",
       "      <td>393</td>\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>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>394</td>\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>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>395</td>\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>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>396</td>\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>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>397</td>\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>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      MPG  Cylinders  Displacement  Horsepower  Weight  Acceleration  \\\n",
       "393  27.0          4         140.0        86.0  2790.0          15.6   \n",
       "394  44.0          4          97.0        52.0  2130.0          24.6   \n",
       "395  32.0          4         135.0        84.0  2295.0          11.6   \n",
       "396  28.0          4         120.0        79.0  2625.0          18.6   \n",
       "397  31.0          4         119.0        82.0  2720.0          19.4   \n",
       "\n",
       "     Model Year  USA  Europe  Japan  \n",
       "393          82  1.0     0.0    0.0  \n",
       "394          82  0.0     1.0    0.0  \n",
       "395          82  1.0     0.0    0.0  \n",
       "396          82  1.0     0.0    0.0  \n",
       "397          82  1.0     0.0    0.0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset['USA'] = (origin == 1) * 1.0  # Cast Series to float\n",
    "dataset['Europe'] = (origin == 2) * 1.0\n",
    "dataset['Japan'] = (origin == 3) * 1.0\n",
    "dataset.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Split the data set into an 80/20 train/test division."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataset = dataset.sample(frac=0.8, random_state=0)\n",
    "test_dataset = dataset.drop(train_dataset.index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x1516fa978>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 20 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.pairplot(train_dataset[[\"MPG\", \"Cylinders\", \"Displacement\", \"Weight\"]], diag_kind=\"kde\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "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",
       "      <td>Cylinders</td>\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",
       "      <td>Displacement</td>\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",
       "      <td>Horsepower</td>\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",
       "      <td>Weight</td>\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",
       "      <td>Acceleration</td>\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",
       "      <td>Model Year</td>\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",
       "      <td>USA</td>\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",
       "      <td>Europe</td>\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",
       "      <td>Japan</td>\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>"
      ],
      "text/plain": [
       "              count         mean         std     min      25%     50%  \\\n",
       "Cylinders     314.0     5.477707    1.699788     3.0     4.00     4.0   \n",
       "Displacement  314.0   195.318471  104.331589    68.0   105.50   151.0   \n",
       "Horsepower    314.0   104.869427   38.096214    46.0    76.25    94.5   \n",
       "Weight        314.0  2990.251592  843.898596  1649.0  2256.50  2822.5   \n",
       "Acceleration  314.0    15.559236    2.789230     8.0    13.80    15.5   \n",
       "Model Year    314.0    75.898089    3.675642    70.0    73.00    76.0   \n",
       "USA           314.0     0.624204    0.485101     0.0     0.00     1.0   \n",
       "Europe        314.0     0.178344    0.383413     0.0     0.00     0.0   \n",
       "Japan         314.0     0.197452    0.398712     0.0     0.00     0.0   \n",
       "\n",
       "                  75%     max  \n",
       "Cylinders        8.00     8.0  \n",
       "Displacement   265.75   455.0  \n",
       "Horsepower     128.00   225.0  \n",
       "Weight        3608.00  5140.0  \n",
       "Acceleration    17.20    24.8  \n",
       "Model Year      79.00    82.0  \n",
       "USA              1.00     1.0  \n",
       "Europe           0.00     1.0  \n",
       "Japan            0.00     1.0  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_stats = train_dataset.describe()\n",
    "train_stats.pop(\"MPG\") \n",
    "train_stats = train_stats.transpose()\n",
    "train_stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_labels = train_dataset.pop('MPG')\n",
    "test_labels = test_dataset.pop('MPG')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now let's normal-transform our training and test data sets for use in a regression network."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Why do we need to do this normal transform?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are broadly two reasons that a normalizing transformation is good practice before training a neural network. First, it ensures that all input features are commensurate with one another and treated equally -- variance in one feature due to units doesn't exact an undue influence over the entire model. Second, it enforces that the network is receiving both positive and negative inputs, which (depending on your activation function) might have an impact on the model's training performance.\n",
    "\n",
    "For more detail on this topic, I recommend the following blog post from Timo Stöttner:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe src=\"https://towardsdatascience.com/why-data-should-be-normalized-before-training-a-neural-network-c626b7f66c7d\" width=\"900\" height=\"500\"></iframe>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%HTML\n",
    "<iframe src=\"https://towardsdatascience.com/why-data-should-be-normalized-before-training-a-neural-network-c626b7f66c7d\" width=\"900\" height=\"500\"></iframe>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def norm(x: pd.DataFrame) -> pd.DataFrame:\n",
    "    \"\"\"Norm all columns to zero mean and unit standard deviation.\"\"\"\n",
    "    return (x - train_stats['mean']) / train_stats['std']\n",
    "\n",
    "normed_train_data = norm(train_dataset)\n",
    "normed_test_data = norm(test_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_model() -> tf.keras.Model:\n",
    "    model = keras.Sequential([\n",
    "        layers.Dense(64, activation='relu', input_shape=[len(train_dataset.keys())]),\n",
    "        layers.Dense(64, activation='relu'),\n",
    "        layers.Dense(1),\n",
    "    ])\n",
    "\n",
    "    optimizer = tf.keras.optimizers.RMSprop(0.001)\n",
    "\n",
    "    model.compile(loss='mse',\n",
    "                optimizer=optimizer,\n",
    "                metrics=['mae', 'mse'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = build_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 64)                640       \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 64)                4160      \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 1)                 65        \n",
      "=================================================================\n",
      "Total params: 4,865\n",
      "Trainable params: 4,865\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4,865 parameters -- how many observations do we have, anyway?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(314, 9)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "normed_train_data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pass an example batch through the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.12717636],\n",
       "       [0.23551516],\n",
       "       [0.42345396],\n",
       "       [0.2064361 ],\n",
       "       [0.30734527],\n",
       "       [0.21771821],\n",
       "       [0.27875608],\n",
       "       [0.03437432],\n",
       "       [0.21414956],\n",
       "       [0.11211875]], dtype=float32)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "example_batch = normed_train_data[:10]\n",
    "example_result = model.predict(example_batch)\n",
    "example_result"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And again, just for good measure."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.12717636],\n",
       "       [0.23551516],\n",
       "       [0.42345396],\n",
       "       [0.2064361 ],\n",
       "       [0.30734527],\n",
       "       [0.21771821],\n",
       "       [0.27875608],\n",
       "       [0.03437432],\n",
       "       [0.21414956],\n",
       "       [0.11211875]], dtype=float32)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(example_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that we got the **same** predictions when passing the same training records through the graph."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What will happen if we train a _new_ model and predict on these same records?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.2932623 ],\n",
       "       [-0.07601981],\n",
       "       [-0.38309133],\n",
       "       [-0.2754855 ],\n",
       "       [-0.17604262],\n",
       "       [-0.02785384],\n",
       "       [-0.17826444],\n",
       "       [-0.0513538 ],\n",
       "       [-0.02715447],\n",
       "       [-0.0671993 ]], dtype=float32)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model2 = build_model()\n",
    "\n",
    "model2.predict(example_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Different results. That's due to _random initialization of the network weights_. But, of course, these networks are untrained, so no real concern."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll use a `keras` _callback_ to print a dot at the end of every epoch, just to give us a sense of how quickly things are progressing. We'll train for 1,000 epochs using an 80/20 _training/validation_ split."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "...................................................................................................."
     ]
    }
   ],
   "source": [
    "class PrintDot(keras.callbacks.Callback):\n",
    "    \"\"\"\n",
    "    Display training progress by printing a single dot for \n",
    "    each completed epoch.\n",
    "    \"\"\"\n",
    "    def on_epoch_end(self, epoch, logs) -> None:\n",
    "        \"\"\"\n",
    "        `on_epoc_end` hook to take the epoch number and logs.\n",
    "        This will run whenever the training loop finishes an epoch.\n",
    "        \"\"\"\n",
    "        if epoch % 100 == 0: \n",
    "            print('')\n",
    "        print('.', end='')\n",
    "\n",
    "        \n",
    "EPOCHS = 1000\n",
    "\n",
    "\n",
    "history = model.fit(\n",
    "    normed_train_data, \n",
    "    train_labels,\n",
    "    epochs=EPOCHS, \n",
    "    validation_split = 0.2, \n",
    "    verbose=0,\n",
    "    callbacks=[PrintDot()]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Because we've outputted the training results to the `history` variable, we can access the results of each epoch natively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x1561b57b8>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What methods are available on this object?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['__class__',\n",
       " '__delattr__',\n",
       " '__dict__',\n",
       " '__dir__',\n",
       " '__doc__',\n",
       " '__eq__',\n",
       " '__format__',\n",
       " '__ge__',\n",
       " '__getattribute__',\n",
       " '__gt__',\n",
       " '__hash__',\n",
       " '__init__',\n",
       " '__init_subclass__',\n",
       " '__le__',\n",
       " '__lt__',\n",
       " '__module__',\n",
       " '__ne__',\n",
       " '__new__',\n",
       " '__reduce__',\n",
       " '__reduce_ex__',\n",
       " '__repr__',\n",
       " '__setattr__',\n",
       " '__sizeof__',\n",
       " '__str__',\n",
       " '__subclasshook__',\n",
       " '__weakref__',\n",
       " '_chief_worker_only',\n",
       " '_keras_api_names',\n",
       " '_keras_api_names_v1',\n",
       " 'epoch',\n",
       " 'history',\n",
       " 'model',\n",
       " 'on_batch_begin',\n",
       " 'on_batch_end',\n",
       " 'on_epoch_begin',\n",
       " 'on_epoch_end',\n",
       " 'on_predict_batch_begin',\n",
       " 'on_predict_batch_end',\n",
       " 'on_predict_begin',\n",
       " 'on_predict_end',\n",
       " 'on_test_batch_begin',\n",
       " 'on_test_batch_end',\n",
       " 'on_test_begin',\n",
       " 'on_test_end',\n",
       " 'on_train_batch_begin',\n",
       " 'on_train_batch_end',\n",
       " 'on_train_begin',\n",
       " 'on_train_end',\n",
       " 'params',\n",
       " 'set_model',\n",
       " 'set_params',\n",
       " 'validation_data']"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dir(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we'll use the `history.history` dictionary accessor and cast it to a `pandas` `DataFrame`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "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>loss</th>\n",
       "      <th>mae</th>\n",
       "      <th>mse</th>\n",
       "      <th>val_loss</th>\n",
       "      <th>val_mae</th>\n",
       "      <th>val_mse</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>558.313916</td>\n",
       "      <td>22.292562</td>\n",
       "      <td>558.313904</td>\n",
       "      <td>551.682411</td>\n",
       "      <td>22.039917</td>\n",
       "      <td>551.682434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>509.329371</td>\n",
       "      <td>21.103104</td>\n",
       "      <td>509.329407</td>\n",
       "      <td>501.714747</td>\n",
       "      <td>20.801769</td>\n",
       "      <td>501.714752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>460.376780</td>\n",
       "      <td>19.866629</td>\n",
       "      <td>460.376770</td>\n",
       "      <td>445.857694</td>\n",
       "      <td>19.382477</td>\n",
       "      <td>445.857697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>406.390395</td>\n",
       "      <td>18.470711</td>\n",
       "      <td>406.390411</td>\n",
       "      <td>385.455736</td>\n",
       "      <td>17.754887</td>\n",
       "      <td>385.455719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>348.474478</td>\n",
       "      <td>16.855032</td>\n",
       "      <td>348.474518</td>\n",
       "      <td>320.984907</td>\n",
       "      <td>15.973777</td>\n",
       "      <td>320.984894</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         loss        mae         mse    val_loss    val_mae     val_mse\n",
       "0  558.313916  22.292562  558.313904  551.682411  22.039917  551.682434\n",
       "1  509.329371  21.103104  509.329407  501.714747  20.801769  501.714752\n",
       "2  460.376780  19.866629  460.376770  445.857694  19.382477  445.857697\n",
       "3  406.390395  18.470711  406.390411  385.455736  17.754887  385.455719\n",
       "4  348.474478  16.855032  348.474518  320.984907  15.973777  320.984894"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hist_df = pd.DataFrame(history.history)\n",
    "hist_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, that gives us the loss, mean absolute error, and mean squared error from both the training and validation sets, indexed by epoch. To make this explicit, let's tack on the epoch indices as a column themselves."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "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>loss</th>\n",
       "      <th>mae</th>\n",
       "      <th>mse</th>\n",
       "      <th>val_loss</th>\n",
       "      <th>val_mae</th>\n",
       "      <th>val_mse</th>\n",
       "      <th>epoch</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>995</td>\n",
       "      <td>2.428237</td>\n",
       "      <td>1.019923</td>\n",
       "      <td>2.428237</td>\n",
       "      <td>7.194723</td>\n",
       "      <td>2.053447</td>\n",
       "      <td>7.194724</td>\n",
       "      <td>995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>996</td>\n",
       "      <td>2.428793</td>\n",
       "      <td>1.048187</td>\n",
       "      <td>2.428793</td>\n",
       "      <td>7.258256</td>\n",
       "      <td>2.113956</td>\n",
       "      <td>7.258256</td>\n",
       "      <td>996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>997</td>\n",
       "      <td>2.381019</td>\n",
       "      <td>0.994397</td>\n",
       "      <td>2.381019</td>\n",
       "      <td>7.278555</td>\n",
       "      <td>2.090306</td>\n",
       "      <td>7.278555</td>\n",
       "      <td>997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>998</td>\n",
       "      <td>2.379971</td>\n",
       "      <td>1.019035</td>\n",
       "      <td>2.379970</td>\n",
       "      <td>7.139082</td>\n",
       "      <td>2.036724</td>\n",
       "      <td>7.139082</td>\n",
       "      <td>998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>999</td>\n",
       "      <td>2.411838</td>\n",
       "      <td>1.008455</td>\n",
       "      <td>2.411838</td>\n",
       "      <td>7.058106</td>\n",
       "      <td>2.054368</td>\n",
       "      <td>7.058106</td>\n",
       "      <td>999</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         loss       mae       mse  val_loss   val_mae   val_mse  epoch\n",
       "995  2.428237  1.019923  2.428237  7.194723  2.053447  7.194724    995\n",
       "996  2.428793  1.048187  2.428793  7.258256  2.113956  7.258256    996\n",
       "997  2.381019  0.994397  2.381019  7.278555  2.090306  7.278555    997\n",
       "998  2.379971  1.019035  2.379970  7.139082  2.036724  7.139082    998\n",
       "999  2.411838  1.008455  2.411838  7.058106  2.054368  7.058106    999"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hist_df['epoch'] = history.epoch\n",
    "hist_df.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Seeing as we've got both the training and validation stats by epoch, it makes natural sense to plot them and see how they evolve as the network trains."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_history(history: tf.keras.callbacks.History) -> None:\n",
    "    \"\"\"\n",
    "    Plot training vs validation MAE and MSE over epoch.\n",
    "    \"\"\"\n",
    "    hist = pd.DataFrame(history.history)\n",
    "    hist['epoch'] = history.epoch\n",
    "\n",
    "    plt.figure()\n",
    "    plt.xlabel('Epoch')\n",
    "    plt.ylabel('Mean Abs Error [MPG]')\n",
    "    plt.plot(hist['epoch'], hist['mae'],\n",
    "           label='Train Error')\n",
    "    plt.plot(hist['epoch'], hist['val_mae'],\n",
    "           label = 'Val Error')\n",
    "    plt.ylim([0,5])\n",
    "    plt.legend()\n",
    "\n",
    "    plt.figure()\n",
    "    plt.xlabel('Epoch')\n",
    "    plt.ylabel('Mean Square Error [$MPG^2$]')\n",
    "    plt.plot(hist['epoch'], hist['mse'],\n",
    "           label='Train Error')\n",
    "    plt.plot(hist['epoch'], hist['val_mse'],\n",
    "           label = 'Val Error')\n",
    "    plt.ylim([0,20])\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "plot_history(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That's not good. It looks like our validation error stopped improving after ~200 epochs and actually got _worse_ around 500-600 epochs. This seems like textbook [overfitting](https://en.wikipedia.org/wiki/Overfitting)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe src=\"https://en.wikipedia.org/wiki/Overfitting\" width=\"900\" height=\"500\"></iframe>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%HTML\n",
    "<iframe src=\"https://en.wikipedia.org/wiki/Overfitting\" width=\"900\" height=\"500\"></iframe>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How about if we predict on our example batch again?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[26.87983 ],\n",
       "       [23.134052],\n",
       "       [12.875172],\n",
       "       [37.10966 ],\n",
       "       [35.73888 ],\n",
       "       [19.928654],\n",
       "       [38.01875 ],\n",
       "       [30.57543 ],\n",
       "       [20.586237],\n",
       "       [21.513508]], dtype=float32)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(example_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And are those predictions stable/deterministic?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[26.87983 ],\n",
       "       [23.134052],\n",
       "       [12.875172],\n",
       "       [37.10966 ],\n",
       "       [35.73888 ],\n",
       "       [19.928654],\n",
       "       [38.01875 ],\n",
       "       [30.57543 ],\n",
       "       [20.586237],\n",
       "       [21.513508]], dtype=float32)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(example_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And how does that compare to the completely untrained model?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.2932623 ],\n",
       "       [-0.07601981],\n",
       "       [-0.38309133],\n",
       "       [-0.2754855 ],\n",
       "       [-0.17604262],\n",
       "       [-0.02785384],\n",
       "       [-0.17826444],\n",
       "       [-0.0513538 ],\n",
       "       [-0.02715447],\n",
       "       [-0.0671993 ]], dtype=float32)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model2.predict(example_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How about if we were to train that network? Would we get the same results for the same input?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "...................................................................................................."
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x156c5d160>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model2.fit(\n",
    "    normed_train_data, \n",
    "    train_labels,\n",
    "    epochs=EPOCHS, \n",
    "    validation_split = 0.2, \n",
    "    verbose=0,\n",
    "    callbacks=[PrintDot()]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(array([26.87983], dtype=float32), array([27.628386], dtype=float32)),\n",
       " (array([23.134052], dtype=float32), array([23.001629], dtype=float32)),\n",
       " (array([12.875172], dtype=float32), array([12.779822], dtype=float32)),\n",
       " (array([37.10966], dtype=float32), array([37.14297], dtype=float32)),\n",
       " (array([35.73888], dtype=float32), array([35.638794], dtype=float32)),\n",
       " (array([19.928654], dtype=float32), array([19.49616], dtype=float32)),\n",
       " (array([38.01875], dtype=float32), array([37.821285], dtype=float32)),\n",
       " (array([30.57543], dtype=float32), array([30.162401], dtype=float32)),\n",
       " (array([20.586237], dtype=float32), array([20.233015], dtype=float32)),\n",
       " (array([21.513508], dtype=float32), array([20.641964], dtype=float32))]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(zip(model.predict(example_batch), model2.predict(example_batch)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What are the actual differences in prediction value?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.74855614],\n",
       "       [ 0.1324234 ],\n",
       "       [ 0.09534931],\n",
       "       [-0.03330994],\n",
       "       [ 0.10008621],\n",
       "       [ 0.4324932 ],\n",
       "       [ 0.19746399],\n",
       "       [ 0.41302872],\n",
       "       [ 0.3532219 ],\n",
       "       [ 0.8715439 ]], dtype=float32)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(example_batch) - model2.predict(example_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Early stopping"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can prevent overfitting to some degree via _early stopping_, wherein we monitor improvement in some metric and end the training routine after some tolerance of no improvement. `tf.keras` provides this for us as a built-in callback object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe src=\"https://en.wikipedia.org/wiki/Early_stopping\" width=\"900\" height=\"500\"></iframe>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%HTML\n",
    "<iframe src=\"https://en.wikipedia.org/wiki/Early_stopping\" width=\"900\" height=\"500\"></iframe>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's give it a shot on our model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "......................................................"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = build_model()\n",
    "\n",
    "# The patience parameter is the amount of epochs to check for improvement\n",
    "early_stop = keras.callbacks.EarlyStopping(monitor='val_loss', patience=10)\n",
    "# In this example, we'll monitor the validation loss for improvement and tolerate 10 epochs of stagnation.\n",
    "\n",
    "history = model.fit(\n",
    "    normed_train_data, \n",
    "    train_labels, \n",
    "    epochs=EPOCHS,\n",
    "    validation_split = 0.2, \n",
    "    verbose=0, \n",
    "    callbacks=[early_stop, PrintDot()]\n",
    ")\n",
    "\n",
    "plot_history(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We got through less than 100 epochs before pulling the plug due to no improvement in the validation loss function!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's try again on a second model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "............................................"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAEKCAYAAAAYd05sAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAAgAElEQVR4nO3deXxU5dn4/8+VyYRsJCELEEggBER2AkQWRQXcK+BeBbRW64PaTe3Tx60+P5dutv224tLWpeLKA61WrdUqbgjumCCLguwBAiFkIyH7dv3+OBMIkI1kJpNMrvfrdV4z58xZrhzlmnvucy+iqhhjjAk8Qf4OwBhjjG9YgjfGmABlCd4YYwKUJXhjjAlQluCNMSZAWYI3xpgAFezLk4tIFnAIqANqVTXdl9czxhhzhE8TvMdMVc3vhOsYY4xpxKpojDEmQIkve7KKyE6gCFDgCVV9sol9FgILASIiIiaNGDHCZ/EEhLpqyP0GopPZVRlOdW09J/WL9HdUxhg/yczMzFfVhKY+83WCH6iqe0WkL/Au8BNVXdXc/unp6ZqRkeGzeAKCKvxhKJx8AfcH/ZB/fLmHr+8/DxHxd2TGGD8Qkczmnm/6tIpGVfd6Xg8ArwKTfXm9HkEEEtMgZx1JfcIpq66jqLzG31EZY7ognyV4EYkQkd4N74Fzga99db0eJXE8HNjE4CjnP192UbmfAzLGdEW+LMH3Az4WkXXAauBNVX3bh9frOQakQX0tqfW7ANhTWOHngIwxXZHPmkmq6g5gvK/O36MlpjkvFZuBAVaCN11CTU0N2dnZVFZW+juUgBQaGkpSUhJut7vNx3RGO3jjbTGDIDSGsLwNRIcNZo8leNMFZGdn07t3b1JSUuyhv5epKgUFBWRnZzNkyJA2H2ft4LsjEaeaJmctSX3CyC6yKhrjf5WVlcTFxVly9wERIS4u7oR/HVmC764Sx0PuRlJi3JbgTZdhyd132nNvLcF3V4lpUF/DhF45ZBeVY1MvGmOOZQm+uxrgPGgdyXYqa+rJL632c0DG+FdBQQFpaWmkpaXRv39/Bg4ceHi9urpt/z6uu+46Nm/e3OZr/u1vfyMhIeHwddLS0k7oeF+zh6zdVZ8h0CuawdXbgJHsLiwjoXcvf0dljN/ExcWxdu1aAO677z4iIyP5+c9/ftQ+qoqqEhTUdNn2mWeeOeHrLliwgEWLFjX7eW1tLcHBR1JtazE0VldXh8vlOuGYGlgJvrsSgcRx9C3dBMC6PcV+DsiYrmnbtm2MGjWKBQsWMHr0aHJycli4cCHp6emMHj2aBx544PC+06dPZ+3atdTW1hITE8Odd97J+PHjmTZtGgcOHGjzNd977z1mzJjB7NmzGTt2bJMxvPjii4wdO5YxY8Zw9913Axy+7q233sq4ceNYvXp1h/52K8F3Z4njCVn9FMlRwWTuLuJ62t58yhhfuv/f37BxX4lXzzlqQBT3zhndrmO//fZbnn/+edLTnSFbHnzwQWJjY6mtrWXmzJlcfvnljBo16qhjiouLOfPMM3nwwQf52c9+xuLFi7nzzjuPO/eSJUv48MMPD683JOWMjAw2btzIoEGD2LZt21ExZGdnc88995CRkUF0dDRnn302b7zxBueffz7FxcWcccYZLf4qaCsrwXdnAyZAXRUXJpaQmVVkD1qNacbQoUMPJ3eApUuXMnHiRCZOnMimTZvYuHHjcceEhYVxwQUXADBp0iSysrKaPPeCBQtYu3bt4SUkJASAadOmMWjQoCZj+OKLL5g1axbx8fG43W7mz5/PqlXOOIwhISFccsklXvm7rQTfnXl6tE6PyObxknD2HqwgqU+4n4MyhnaXtH0lIiLi8PutW7fy8MMPs3r1amJiYrj66qubbF/ekKgBXC4XtbW17b5mU+vNCQsL81pzUyvBd2exqRDSm5HsACBzV5GfAzKm6yspKaF3795ERUWRk5PD8uXLOz2GKVOmsGLFCgoKCqitrWXZsmWceeaZXr+OleC7s6AgSBxHbPEmwkMuJCOriIvSBvo7KmO6tIkTJzJq1ChGjBjB4MGDOe200zp0vmPr4J944olWj0lKSuKXv/wlM2bMQFWZM2cOF1544Qn/SmiNTyf8OFE24Uc7vH0XZDzDNf3+SUF5Pf+55XR/R2R6qE2bNjFy5Eh/hxHQmrrHfpvww3SCxDSoreCchIN8u7+E0irvlgCMMd2XJfjubvA0AKazlnqFr3ZbPbwxxmEJvruLGQSJ4xl04H1EICPLErwxxmEJPhCMnEPwvkxOTai2ljTGmMMswQeCkXMBuKr3Or7aXURtXb2fAzLGdAWW4ANBwskQP5ypVZ9SVl3H5txD/o7IGNMFWIIPFCPnEF+QQQyHrJrG9EgzZ848rtPSokWLuPnmm1s8LjIyssntLpfrqGGAH3zwQa/F2lkswQeKEbMRrePSiPX2oNX0SPPmzWPZsmVHbVu2bBnz5s1r1/nCwsKOGmOmqYHG6urqjlpva0clb3doao4l+EAxYAJEJXFx6BorwZse6fLLL+fNN988PLlHVlYW+/bt4/TTT6e0tJSzzjqLiRMnMnbsWP71r3+1+zopKSnccccdTJw4kZdeeokZM2Zw6623kp6ezsMPP0xWVhazZs1i3LhxnHXWWezevRuA73//+9x0001MmTKF22+/3St/c2tsqIJAIQIj5zDqy6c5WF5ITnEFidFh/o7K9FRv3Qn7N3j3nP3HwgXNV5PExsYyefJk3nrrLS666CKWLVvGd7/7XUSE0NBQXn31VaKiosjPz2fq1KnMnTu3xUG9KioqSEtLO7x+1113ceWVVwLO5CJr1qwB4PHHH6e6upqGXvhz5szh2muv5dprr2Xx4sX89Kc/5bXXXgMgOzubTz/9tEOTeJwIK8EHkpFzCK6v5sygdVaKNz1S42qaxtUzqsrdd9/NuHHjOPvss9m7dy+5ubktnuvYKpqG5A4c9f7Y9c8++4z58+cDcM011/Dxxx8f/uyKK67otOQOVoIPLIOmouHxXFifwZdZRcweN8DfEZmeqoWSti9ddNFF3HbbbaxZs4by8nImTZoEOAOC5eXlkZmZidvtJiUlpckhgtuqvUMBt3U/b7ESfCAJciEjvsMs11esz2q5dGJMIIqMjGTmzJlcf/31Rz1cLS4upm/fvrjdblasWMGuXbt8FsOpp556+FfEkiVLOP10/w0AaAk+0IycS5hW0Cf3U8ps4DHTA82bN49169YdleAXLFhARkYGY8eO5fnnn2fEiBGtnqehDr5haaoVTVMeffRRnnnmGcaNG8cLL7zAww8/3O6/paNsuOBAU1tF7e9SebkinUHff5pTh8X7OyLTQ9hwwb5nwwX3dMG9qB92Hue4MlmTlefvaIwxfmQJPgCFjJlLnBzi0JaPW9/ZGBOwLMEHomFnUyMhDDrwHvX1XacKzgS+rlTlG2jac28twQeiXpHk9Z3OTF3NltwSf0djeojQ0FAKCgosyfuAqlJQUEBoaOgJHWft4AOUe+xFJOR+wLr1HzMi8UJ/h2N6gKSkJLKzs8nLs2c/vhAaGkpSUtIJHWMJPkDFT5xL7Xu34d7yJpxnCd74ntvtZsiQIf4OwzRiVTQBSsJj2RKWxvDCFWA/mY3pkXye4EXEJSJficgbvr6WOVph8nkM0n0UZK3zdyjGGD/ojBL8LcCmTriOOUb0pEup0mBKVv3V36EYY/zApwleRJKAC4G/+fI6pmmjTjqJt1wzSdr5Tyg94O9wjDGdzNcl+EXA7UCzs0CLyEIRyRCRDHv67l2uIOHAuBsJ0lrKVj7i73CMMZ3MZwleRGYDB1Q1s6X9VPVJVU1X1fSEhARfhdNjzTp1Gm/VT8G95hmoLPZ3OMaYTuTLEvxpwFwRyQKWAbNE5EUfXs80YVjfSN6LW0BIXSn65dP+DscY04l8luBV9S5VTVLVFOAq4ANVvdpX1zPNmzTlDFbWjaP2079ATYW/wzHGdBJrB98DzBk/gCf1YtwVebB2ib/DMcZ0kk5J8Kr6oarO7oxrmePFhIcQdfKZrGc4+skjUGcTgRjTE1gJvoe4bFIyj1TPQQ7ugm9e9Xc4xphOYAm+hzjz5ATWhk5hX8hg+PghG77AmB7AEnwP4XYFMWdCEg+VXwgHvoGt7/g7JGOMj1mC70Eum5jEq7VTKQ0dAB/9yd/hGGN8zBJ8DzJ6QBTD+vdhSfBFsOdz2PWZv0MyxviQJfgeRES4dOJAHsqfTG1YHHxspXhjAlmLCV5EYtuwxHRWsKbjLk4bSLX04tP4K5x6+H1r/R2SMcZHWpvRaZ9nkRb2cQGDvBaR8am+UaGcMTyBX+WcyvLwl5F/3gD/9T6ERvs7NGOMl7VWRbNJVVNVdUhzC1DQGYEa77lsYhJbSoLZcOojULQTXrkR6psd8NMY0021luCnteEcbdnHdCHnjOpH79Bgnt07EM77DWx5C1b+zt9hGWO8rMUEr6qVjddFZKCIDPIswU3tY7q+ULeL2eMSeevr/ZSOvx7Gz4eVD8K3b/o7NGOMF7X2kPUuEfn/Gm36DHgDeAf4H18GZnzrsolJVNTU8fq6HJj9EAyY4FTV5G3xd2jGGC9prYrmCuCPjdYLVHUcMBpnKj7TTU0a3IfxSdH8ecU2qiUErnwR3KGwbJ5NDGJMgGi1HbyqljVafdizrQ4I81VQxvdEhFvPGc7egxW8lLkHopPgiuegKAteWWgPXY0JAK0l+EgRcTesqOqzACLSC4jyYVymE8wYnsCEQTE89sE2qmrrIOU0OO+3sOVtp07eGNOttZbgXwaeEJHwhg0iEgE87vnMdGMiws/OGU5OcSX/+HKPs3Hyf0HaAqdVzea3/RugMaZDWkvw/wscAHaLSKaIrAGygFzPZ6abmz4snlNS+vDYim1U1tSBCFz4J+g3Bv59C1Qc9HeIxph2aq2ZZJ2q3gkkA98HrgUGqeqdqmrTAgUAEeG2c4aTW1LF0tW7nY3uUJj7KJQdgPfu9W+Axph2a62Z5Eki8i/gS+BuoFBVbdbmAHPq0Himpsbylw+3U1Fd52wcOBGm/hAyn4Wsj/0anzGmfVqrolmM0+79MmAN8KjPIzJ+cdvZw8k7VMWSL3Yd2TjzF9AnBV7/KdTY97ox3U1rCb63qj6lqptV9Q9ASifEZPxgSmoc04fF89cPt1Ne7al9CwmH2YugcDus/L1/AzTGnLDWEnyoiEwQkYkiMhEIO2bdBJDbzjmJgrJqnv+sUSl+6ExIuxo+eRhy1vsvOGPMCRNtYfJlEfkQaG4HVdVZ3gwmPT1dMzIyvHlKc4KuXbya9dkH+eiOWUT28owmXV4If54CUQPghvfB1doo08aYziIimaqa3tRnrbWimaGqM5tZvJrcTddw2znDKSqv4blPs45sDI+F7/wectbC53/xW2zGmBPTYlFMRC5t6XNVfcW74Rh/S0uO4awRfXly1Q6umTaYqFBPR+ZRF8PJ34EVv4GRsyE21b+BGmNa1ZaerPcAsz3LnEbLbN+GZvzltnOGU1xRw6J3tx7ZKAIX/hFcbqcDVAtVe8aYrqG1BH8psAUYB+wEfq2q13mW630enfGLMQOj+d60wSz+ZCcffJt75IOoAXD2fbBzFXz2mL/CM8a0UWt18K+p6lXAmcB24I8i8rGInNkp0Rm/ufs7IxmVGMV//2MdOcWN2sBPus6pqnnnHlj+C6iv81+QxpgWtTpcsEclUAyUAJFAqM8iMl1CqNvFY/MnUFVbzy1L11Jb5xk+OCgIvvsCTF7olOL/fg1Ul7V8MmOMX7Q2VMEsEXkSyARmAg+rapqqLu+U6IxfpSZE8utLxrA6q5BH3m9UH+8Khu/8AS74vTOf6zMXQMk+/wVqjGlSayX494DJwMdAL+B7IvJIw+Lz6IzfXTIhiSsmJfHoim18ui3/6A+n3Ajz/g4F2+GpsyBnnX+CNMY0qbUEfz3wEM5gYxk4JfnGi+kB7r9oNEMTIrnl72vJO1R19IfDz4Xrl4MEweILYPNb/gnSGHOcFnuydjbrydp1fbu/hIse+4TJQ2J57rrJBAXJ0Tsc2g9L58G+r+CC3zmle2OMz7W7J6uI3NeGk7e6j+n+RvSP4t45o/loaz6Pr9p+/A69+8P334QRF8Jbt8OHD1pbeWP8rLVBRW4QkZIWPhfgKuC+4z4QCQVW4dTdBwMvq6rNHtGNzZuczCfb8/njO1tIHxzL5CGxR+8QEu5M3P3vn8KHv4XKYjj3107LG2NMp2vtX95TQO8WlkjPPk2pAmap6nggDThfRKZ6I2jjHyLCby8dy6DYcK5/9ku+zCo8fidXMMx9zJks5PO/wOs/hjqb/MsYf2ixBK+q97f3xOpU7pd6Vt2exX6zd3NRoW6W/tdU5j/1Od97ejVPX5vOqcPij94pKAjO+w2ExsCHv4GqErjsaQju5Z+gjemhfPrbWURcIrIWZ+Lud1X1iyb2WSgiGSKSkZeX58twjJf0jw5l2Y1TSY4N47pnv+TDzQeO30kEZtwB5z8Im/4N/3eldYgyppP5NMF7Ju1OA5KAySIypol9nlTVdFVNT0hI8GU4xov69g5l2cJpDOsbycLnM3nnm/1N7zj1Zrj4r7BzJTx/MVQUdW6gxvRgrSZ4Tyn8to5cRFUPAiuA8ztyHtO1xEaE8H83TGXUgCh+uGQNb67PaXrHtPnOw9ectfDiZTa/qzGdpNUEr6p1wLwTPbGIJIhIjOd9GHAO8O0JR2i6tOhwNy/8YDITBsXwk6VrePWr7KZ3HDUXLn8G9q6B135oTSiN6QRtraL5REQeE5HTG+ZjbcOcrInAChFZj9MT9l1VfaND0ZouqXeom+eun8zU1Dh+9o91/P3L3U3vOHI2nH0vfPOKTeJtTCdo6+SaaZ7XBxptU6DZaftUdT0woZ1xmW4mPCSYxd8/hZtezOSOf26guk65Zurg43c87VbI2+K0rok/Cca0OGmYMaYD2pTgVXWmrwMx3V+o28UT10ziR0vW8L+vfU1NbT3XTx9y9E4iMGcRFO6A126GPoNh4CT/BGxMgGtTFY2IRIvInxqaM4rIH0Uk2tfBme6nV7CLvyyYxAVj+vPAGxt5YmUTwxoE94KrlkBkX1g634YaNsZH2loHvxg4BHzXs5QAz/gqKNO9hQQH8ei8CcwZP4DfvvUtjzYeS75BRLwz1HB1KSy9ytrIG+MDbU3wQ1X1XlXd4VnuB1J9GZjp3oJdQSy6Mo1LJwzkj+9u4U/vbOa4kUv7jYLLF0POenj1Jqiv90+wxgSotib4ChGZ3rAiIqcB1pjZtMgVJPzhivFcmZ7MIx9s43dvN5Hkh58H5/4KNr0O79gcr8Z4U1tb0dwEPN+o3r0IuNY3IZlA4gpyBigLdgmPr9xOYVkV/zt7FL1D3Ud2mvYjOLjLGZwsZx1c+iREJ/kvaGMCRFt6sgYBJ3tGhRwHjFPVCZ5mkMa0KihI+NXFY/jhjKG8lJnNOX9adfTQBiLOHK+XPOEk+L+e5oxfY4zpkLb0ZK0Hbve8L1HVlsaHN6ZJIsLt54/g1R+eRky4m4UvZHLzi5nkllQe2Wn8VXDjKuiTAn+/Gt64zYY1MKYD2loH/56I/FxEkkUktmHxaWQmIKUlx/Dvn0zn9vNP5oNvD3D2n1ay5Itd1Nd76ubjhsIP3oVTfwoZi+HJmZC70b9BG9NNtWlOVhHZ2cRmVVWvtqSxOVl7lp35Zfzi1Q18ur2AU1L68NtLxzGsb+SRHba977SuqSx25nlNv85/wRrTRbU0J2urCd5TBz9NVT/xRXCNWYLveVSVlzOz+dWbm6ivV/52bTpTUuOO7FCaB6/eCNvfh+m3wVn3OnX2xhigA5Nuw+E6+Me8HpUxOHXzV6Qn89Ytp9M3qhffW7yadzfmHtkhMgEWvASTroOPH3JGoqyr8V/AxnQjba2Df19ELhOxopPxjQExYbx006mM6N+bm17M5J+ZjYYdDnLB7Idgxl2w7v9g2Xzr+WpMG7Q1wd8IvARUiUiJiBwSEWtNY7wqNiKEJf81lampsfz3S+v420c7jnwoAjPudBL9tvfgublQVuC/YI3pBtqU4FW1t6oGqWqIqkZ51qN8HZzpeSJ7OcMOnz+6P796cxP/b/kxvV/Tr4fvvgD7N8Di86Bol/+CNaaLazHBi8jVjd6fdsxnP/ZVUKZn6xXs4s8LJnLVKck8tmIb97z2NXX1jZL8yNnwvX9B2QF4+lxnLBtjzHFaK8H/rNH7R4/57Hovx2LMYQ1DHNx05lCWfLGbG1/IJKe4UaenwdPgurdBguCJ0+Gv02H5L2Dre1Y/b4xHi80kReQrVZ1w7Pum1r3Bmkmapjz98U5+9/a3BAncfOYwFp6RSliIy/nw0H746gXYsRL2fAF11RDkhuTJkDoDhp4FAyda00oTsNrdDl5E1qjqxGPfN7XuDZbgTXP2FJbz4Fvf8uaGHAZEh3Lnd0YyZ1wiRzXsqi6HPZ/Djg+dJWc9oJAwEiZ9H8ZfCWF9/PMHGOMjHUnw5cA2QIChnvd41lNVNcKbgVqCN635YkcB9/97IxtzSkgf3Id754xmbFIzk4uVFzqDlmU+C/vWQHAojLrY6RGbPMVK9SYgdCTBNzFr8hGq6tUmDJbgTVvU1SsvZezh/72zmYKyauaOH8DFaQM5bVg8IcHNPFbKWQeZz8H6f0D1IUgYARO/Byd/B2KHNH2MMd1Ah4Yq6EyW4M2JKKms4bEPtrH0i90cqqqld2gwZ4/sxwVj+nPG8ARC3a7jD6oqhW9ecUr1ezOdbfEnw/Bz4aTzYNBUcLmPP86YLsoSvAloVbV1fLItn7c27OedjbkUV9QQHuJi1oi+XJQ2kLNH9qXJTtgF22HrO7BlOWR9DPU10Csahs50lvB4CAkHdwSERBz93h0OQW3tJ2iM71iCNz1GTV09n+8o4D8b9vPON/spKKtmWmocv7x4zNEjVR6r6pDzYHbLctj6LpTub37fBocTfwT0ioSQSAiNhlEXwZjLITjEa3+XMc3xaoIXkT5Asi9mdLIEb7yptq6eZV/u4fdvf0tFTR0Lz0jlxzNPOtLEsjmqUJTlJP3qMqgpc1ro1JRDdamzrbrhfcN6mfO+OBsKd0Bkf5iy0BkkLdymTjC+0+EELyIfAnNx5nDNBA4An6jqz1o67kRZgje+kHeoit/+ZxOvfLWXpD5hPHDRaGaN6Oebi6nC9g/g00dhxwqnKmfC1TD1Zoj16vQJ7VdXA9kZENzL6SPgDflb4Z17YN9XcPp/wyk3OIPEdVRZAaxfBgPTnb4N1vLpON5I8F+p6gQRuQGn9H6viKxX1XHeDNQSvPGlz7YXcM9rG9ieV8Z5o/tx75zRDIgJ890F938Nn/0ZNrwEWgcjLoSkyRDZDyL7HnkNi/V9fX7hDmcCle0rYOcqpyURwNBZMPMXkNRkfmhdRRGs/D2sfhKCw6DfKKfD2YAJMHsRDEhrf8zbVzgTvjRUl8WdBGnzYfw8iEps/3kDjDcS/AbgXOA54Beq+qUleNMdVdfW89RHO3j0g62owgVj+nNFejLTUuMICvJR6bAkx0mAmc84CfFY4nISfb8xzhAMg6bBgIngDm3+nPX1UJIN+Vug4iBoPdTXOV8kjV8PbHImSynKco6LGeT07h06y9n2ySIoL3BaEM28u+0Jua4W1jwLH/za+Zsmfg9m3QMRCfD1P+Htu6A8HybfCLN+Ab16t/1+1VbDBw84v4LiT4a5j0LBVvjqRdj9mTM8xdCznF9GJ1/g/BLpwbyR4K8A/henWuZmEUkF/qCql3kzUEvwprPsKSzniVXbeX3tPkoqaxkYE8Zlk5K4YlISybHhvrmoKlSVQOkBKM31vHreH8pxqjfyvnX2dYU4SX7QVCfh19c6n+VvgbzNzmtNeevXDImElNNhmCepx6YeXc1RVQqrn4BPHoHKgzBitjPufv8xzZ9z+wpYfjcc2AiDp8P5v4XEY8p6FQfh/QeceXV7JzpTLo6c03oVS/5WePl62L/eGTn03F87rZcaFGyHtUtg7VI4tM/pmZw6E4acDilnOHP6dnY1zqFcyN0Aud841V+xQ5z7HJvqPHT3MWtFY0wzKmvqeGdjLi9l7OHjbfmowtTUWC6ZMJCRiVEMjo0gOrwT28WXFThVHLs/hd2fO0m/vvbI51EDIX6401ErYbhTwo2Id34FBAV5Xl1HXsP6tK1df2UJfP5Xp0qpqtipYtF6qK2C2kqoqXRea6ugtgJiBsO5v2o9aWdnwL9vdRJg6kxIPdMT+wjnHA1VU6qw5nl4+06nRD73MWfU0ObU1znPONb/wxmHqKEap3cipEx3vtRSpju/WLzRr0EVyvKch+j5W5zhqnO/dpJ6WV7zx4XHQR9Pwk8Y7tzXxAkQEdf8MSfIGyX4VOBhYCqgwGfAbaq6o8UDT5AleONP+w5W8MqabF7OzCar4EjpODrMzeC4cAbFhjM4LpzBsREkxYaR3CecxOhQgl0+rD+vLneSvDvUSewnUtXRHhVF8NlfnC+Z4FDnusGhTtINDnNeYwbBhGtarkJqrK4WvvgrfP64U63UIDjMSXoJI5wS/9blMOQMuOQJiBrQ9phVnZJ91kfOsvMjZyjpBu4IpyR91BIF7jBw9XJ+LQWHeN67nb+xogiK90LJXiepl+x1BrJr4OoFfUc6v3T6NSyjnXMVZUHRTue5R2Gj1+LdR46PGeQk+4YlMQ3CYtr+NzfijQT/OfBnYKln01XAT1R1SrsiaoYleNMVqCpbD5SyM7+M3QXl7CosY1dBObsKytl7sOKoseldQUJidCjJfcJJ6hPGoNhwRg+MYlxSDPGRPbtuuEmVxU4V04FNzmveJjjwrZNQZ94F037S8QfOqk5Vz+5PnUnbKw861z12afhFUld9ZGkgLudLJmogRA/0vCY5S+xQiBsGruAT/9tz1jlf2A1Lw7ORXlFwx652/e3eSPDHPVAVkXWqOv6Eo2mBJXjT1dXU1bPvYAXZRRXsKSx3XovKD78/cKjq8L4DY8JIS45hfHI045NiGDMwmoheJ5gUegpV/zeBVHWSfG2V03nNG808W1Ne6CT6sjwYf1W7TtFSgm/x/zYRaeih8ZaI3Aksw6miuRL4T7uiMSl93BQAAA6hSURBVKYbc7uCGBwXweC4pgdSLauq5Zt9Jazbc5C12QdZt+cgb27IOfx5SHAQLhGCg4SgIMHlWUJcQcwen8htZw9vegydQOfv5N4QQ3Cvzm2VEx7rPAD3kdZGk9yJk9Cbuvuqql7tuWEleBOI8kurWJ99kK/3llBWXUt9vVJXD3X19dSp8z6/tIp3N+YyJD6C318+jlNSWu/9mneoimWrdxPfuxdXTEry7bMA02X5pBWNiLhVtaaFz5OB54F+OF8ST6rqwy2d0xK86ck+2ZbPHf9cz96DFVw7LYX/Oe/kJqt09h2s4MlVO1i6ejdVtfUAjOjfm/vmjmZqqvdaZ5juwWsJXpwh+WYB84HZqtpsf28RSQQSVXWNiPTGGeLgYlXd2NwxluBNT1dWVcsflm/muc+yGBgTxu8uG8dpw+IByMov468fbueVr7JRhUsmDOTmGUPZvP8Qv3pzE3sPVjBn/ADuumCEb3vomi7FGw9Zp+Ik9YuBWOBHwOuq2kS3vGbP8S/gMVV9t7l9LMEb4/gyq5A7Xl7PjvwyrkxPpqKmjjfW7yPYFcRVpySz8IxUkvoc6QBUUV3H4yu38/jK7QSJ8KOZQ7nh9NTD9fmqSm5JFZtzD7E19xCb9x9CBNIHx3LKkFhS4sKbHlLZdHkdmdHpN8AVwG6cJpKvAhmqekJT4IhICrAKGKOqJcd8thBYCDBo0KBJu3Z5dZIoY7qtypo6HnpvC0+t2kGY28XVUwfzg9OH0Ld38+3P9xSW85v/bOKtr/eTHBvG9GHxbM0tZUvuIUoqj3SYio/sRW19PQfLaw6vn5LSh1NSYpk8JJaT+kUSHBREkNBs4q+rV2rq6qmtV2pq66mprz+qCemx3K4g4iJC7IvEyzqS4A8AW4BFwL9VtUpEdpzIw1URiQRWAr9W1Vda2tdK8MYcL6e4gnB38An1qP1kWz6/fGMjOcWVnNyvN8P7RzK8X+/DS2xECPX1yva8UlZnFZKRVcTqnYXsPVjR5PlcQXIk2SvU1NfTnsd34SEuUuIiGJIQwZC4CIbER5ASH0FqfAR9Irr++Pk5xRWUVdUxOC4cdxd5qN2RBO8CzgHmAWcBK4CzcUaUrG32wCPHu4E3gOWq+qfW9rcEb4x/7TtYwZdZhewuKEdxSumqSr1CvecVwO0SgoOCCHYJbpfgdgUR7AoiOEiabHIHzi+SXYXl7MwvIyu/jD1FR3ca6xPuJjUhkqEJEaQmRJIa77wm9Qkjv7SKPYVOn4PswnL2ePoh5BRXEhsRQlKfMM8Sfvh1YJ8wIr3U7yBzVyFPrdrJ8o37UYXgIGFwXDhDEyIZ1jfy8OuwvpGd3tfBKw9ZRaQXMBsn2Z8OvK+q81vYX3BGnyxU1Vvbcg1L8Mb0HNW19ewpKicrv4wdeWXsyC9je14pO/LKyC+tava4IIHEaCehJ0aHUlReQ3aR09GsoVVRg6jQYBKjw+gfHUpidCiJ0c4x/aNDGRIfwcCYsGZHEa2tq2f5N7n87eMdfLX7INFhbhZMGcTQhEh25Jey7UAp2/OcL6vaRl9UybFhjOgfxYj+vTm5f29G9I8iJS7cZ81Yvd5MUkSicFrEPN/CPtOBj4ANQMNdv1tVm+0gZQneGANQXFHDDk+y33uwgvjIXiTHOkNBJEaHERJ8fLJUVfJLq8kucoaU2FNYwf7iCnKKK9lfUklOcSV5h47+4ghzuxjWN5KT+kVyUt/eDO8XSUp8BCs357H4k51kF1UwOC6cH0wfwuWTkggPOb50XlNXz+7CcrYdKGVr7iG+3e8sO/PLDv9CCQkOIjE6FLcrCLcriBDPrx63Kwh3cBCx4W4WXTWhXffKRpM0xhicXw25JZXsO1jBzvwytuSWsvXAIbbmlrK/pPKofdMH9+GG01M5Z1Q/XO2YK6Cypo7teaVs9iT83JJKauuU6rp6ahqWWme9d2gwL/ygfUN7tXuoAmOMCSQhwUEkx4aTHBvOlGM6hZVU1rA1t5TteaUM79ebtOT2je7YINTtYvSAaEYP8P2Y8M2xBG+MMUBUqJtJg/swaXAff4fiNW1O8CJyKpDS+JiW6uCNMcb4V5sSvIi8AAwF1gJ1ns2KM9aMMcaYLqitJfh0YJR2pSeyxhhjWtTWhplfA/19GYgxxhjvamsJPh7YKCKrgcMNSVV1rk+iMsYY02FtTfD3+TIIY4wx3temBK+qK30diDHGGO9qUx28iEwVkS9FpFREqkWkTkRKWj/SGGOMv7T1IetjOIOMbQXCgBuAP/sqKGOMMR3X5uHNVHUb4FLVOlV9Bjjfd2EZY4zpqLY+ZC0XkRBgrYj8HsjhBL4cjDHGdL62JulrPPv+GCgDkoHLfBWUMcaYjmtrK5pdIhIGJKrq/T6OyRhjjBe0tRXNHJxxaN72rKeJyOu+DMwYY0zHtLWK5j5gMnAQQFXXAkN8FJMxxhgvaGuCr1HV4mO22cBjxhjThbW1Fc03IjIfcInIScBPgU99F5YxxpiOamsJ/ifAaJyBxpYCJcCtvgrKGGNMx7W1FU058AvPYowxphtoMcG31lLGhgs2xpiuq7US/DRgD061zBeA+DwiY4wxXtFagu8PnIMz0Nh84E1gqap+4+vAjDHGdEyLD1k9A4u9rarXAlOBbcCHIvLjTonOGGNMu7X6kFVEegEX4pTiU4BHgFd9G5YxxpiOau0h6/PAGOA/wP2q+nWnRGWMMabDWivBX40zeuQtwE9FDj9jFUBVNcqHsRljjOmAFhO8qtqY78YY001ZAjfGmABlCd4YYwKUJXhjjAlQluCNMSZAWYI3xpgA5bMELyKLReSAiFjbeWOM8QNfluCfBc734fmNMca0wGcJXlVXAYW+Or8xxpiW+b0OXkQWikiGiGTk5eX5OxxjjAkYfk/wqvqkqqaranpCQoK/wzHGmIDh9wRvjDHGNyzBG2NMgPJlM8mlwGfAySKSLSI/8NW1jDHGHK/VCT/aS1Xn+ercxhhjWmdVNMYYE6AswRtjTICyBG+MMQHKErwxxgQoS/DGGBOgLMEbY0yAsgRvjDEByhK8McYEKEvwxhgToCzBG2NMgLIEb4wxAcoSvDHGBChL8MYYE6AswRtjTICyBG+MMQHKErwxxgQoS/DGGBOgLMEbY0yAsgRvjDEByhK8McYEKEvwxhgToCzBG2NMgLIEb4wxAcoSvDHGBChL8MYYE6AswRtjTICyBG+MMQHKErwxxgQoS/DGGBOgLMEbY0yAsgRvjDEByhK8McYEKEvwxhgToCzBG2NMgLIEb4wxAcqnCV5EzheRzSKyTUTu9OW1jDHGHM1nCV5EXMCfgQuAUcA8ERnlq+sZY4w5mi9L8JOBbaq6Q1WrgWXART68njHGmEaCfXjugcCeRuvZwJRjdxKRhcBCz2qpiGxu5/Xigfx2Hhvo7N40ze5L8+zeNK+r3ZvBzX3gywTfJqr6JPBkR88jIhmqmu6FkAKO3Zum2X1pnt2b5nWne+PLKpq9QHKj9STPNmOMMZ3Alwn+S+AkERkiIiHAVcDrPryeMcaYRnxWRaOqtSLyY2A54AIWq+o3vroeXqjmCWB2b5pm96V5dm+a123ujaiqv2MwxhjjA9aT1RhjApQleGOMCVDdPsHbcAhHiMhiETkgIl832hYrIu+KyFbPax9/xugvIpIsIitEZKOIfCMit3i29/j7IyKhIrJaRNZ57s39nu1DROQLz7+tv3saS/Q4IuISka9E5A3Pere5L906wdtwCMd5Fjj/mG13Au+r6knA+571nqgW+G9VHQVMBX7k+X/F7g9UAbNUdTyQBpwvIlOB3wEPqeowoAj4gR9j9KdbgE2N1rvNfenWCR4bDuEoqroKKDxm80XAc573zwEXd2pQXYSq5qjqGs/7Qzj/YAdi9wd1lHpW3Z5FgVnAy57tPfLeiEgScCHwN8+60I3uS3dP8E0NhzDQT7F0Vf1UNcfzfj/Qz5/BdAUikgJMAL7A7g9wuBpiLXAAeBfYDhxU1VrPLj3139Yi4Hag3rMeRze6L909wZsToE6b2B7dLlZEIoF/Areqaknjz3ry/VHVOlVNw+lxPhkY4eeQ/E5EZgMHVDXT37G0l9/HoukgGw6hdbkikqiqOSKSiFNC65FExI2T3Jeo6iuezXZ/GlHVgyKyApgGxIhIsKe02hP/bZ0GzBWR7wChQBTwMN3ovnT3ErwNh9C614FrPe+vBf7lx1j8xlN3+jSwSVX/1OijHn9/RCRBRGI878OAc3CeUawALvfs1uPujarepapJqpqCk1s+UNUFdKP70u17snq+XRdxZDiEX/s5JL8RkaXADJzhTHOBe4HXgH8Ag4BdwHdV9dgHsQFPRKYDHwEbOFKfejdOPXyPvj8iMg7nYaELp9D3D1V9QERScRouxAJfAVerapX/IvUfEZkB/FxVZ3en+9LtE7wxxpimdfcqGmOMMc2wBG+MMQHKErwxxgQoS/DGGBOgLMEbY0yAsgRvehQRqRORtY0Wrw0uJiIpjUfyNMbfuntPVmNOVIWnS74xAc9K8MYAIpIlIr8XkQ2esdGHebaniMgHIrJeRN4XkUGe7f1E5FXPGOrrRORUz6lcIvKUZ1z1dzw9Q43xC0vwpqcJO6aK5spGnxWr6ljgMZze0QCPAs+p6jhgCfCIZ/sjwErPGOoTgYYJ5U8C/qyqo4GDwGU+/nuMaZb1ZDU9ioiUqmpkE9uzcCa92OEZlGy/qsaJSD6QqKo1nu05qhovInlAUuMu6p5hiN/1TB6CiNwBuFX1V77/y4w5npXgjTlCm3l/IhqPSVKHPecyfmQJ3pgjrmz0+pnn/ac4IwkCLMAZsAyc6f1uhsOTZUR3VpDGtJWVLkxPE+aZuajB26ra0FSyj4isxymFz/Ns+wnwjIj8D5AHXOfZfgvwpIj8AKekfjOQgzFdiNXBG8PhOvh0Vc33dyzGeItV0RhjTICyErwxxgQoK8EbY0yAsgRvjDEByhK8McYEKEvwxhgToCzBG2NMgPr/AUnHDou4zLuSAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model2 = build_model()\n",
    "\n",
    "# The patience parameter is the amount of epochs to check for improvement\n",
    "early_stop = keras.callbacks.EarlyStopping(monitor='val_loss', patience=10)\n",
    "\n",
    "history2 = model2.fit(normed_train_data, train_labels, epochs=EPOCHS,\n",
    "                    validation_split = 0.2, verbose=0, callbacks=[early_stop, PrintDot()])\n",
    "\n",
    "plot_history(history2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(54, 44)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(history.epoch), len(history2.epoch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ok, so training the same model different times using the same routine gives us different numbers of required epochs before training terminates."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check out the model predictions between the two models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.99724007],\n",
       "       [ 0.54335594],\n",
       "       [ 0.14667225],\n",
       "       [ 0.26123428],\n",
       "       [ 0.82056046],\n",
       "       [ 0.5086918 ],\n",
       "       [ 1.0368996 ],\n",
       "       [ 0.828207  ],\n",
       "       [ 0.28744507],\n",
       "       [-0.26097298]], dtype=float32)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(example_batch) - model2.predict(example_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Still a difference, although certainly less than before."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "78/78 - 0s - loss: 6.6514 - mae: 2.0151 - mse: 6.6514\n",
      "Testing set Mean Abs Error:  2.02 MPG\n"
     ]
    }
   ],
   "source": [
    "loss1, mae1, mse1 = model.evaluate(normed_test_data, test_labels, verbose=2)\n",
    "\n",
    "print(\"Testing set Mean Abs Error: {:5.2f} MPG\".format(mae1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "78/78 - 0s - loss: 5.8693 - mae: 1.9031 - mse: 5.8693\n",
      "Testing set Mean Abs Error:  1.90 MPG\n"
     ]
    }
   ],
   "source": [
    "loss2, mae2, mse2 = model2.evaluate(normed_test_data, test_labels, verbose=2)\n",
    "\n",
    "print(\"Testing set Mean Abs Error: {:5.2f} MPG\".format(mae2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.055576373"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(mae1 - mae2) / mae1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's plot our predictions vs the true values to see if we're off in any consistent fashion."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_predictions1 = model.predict(normed_test_data).flatten()\n",
    "\n",
    "a = plt.axes(aspect='equal')\n",
    "plt.scatter(test_labels, test_predictions1)\n",
    "plt.xlabel('True Values [MPG]')\n",
    "plt.ylabel('Predictions [MPG]')\n",
    "lims = [0, 50]\n",
    "plt.xlim(lims)\n",
    "plt.ylim(lims)\n",
    "_ = plt.plot(lims, lims)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_predictions2 = model2.predict(normed_test_data).flatten()\n",
    "\n",
    "a = plt.axes(aspect='equal')\n",
    "plt.scatter(test_labels, test_predictions2)\n",
    "plt.xlabel('True Values [MPG]')\n",
    "plt.ylabel('Predictions [MPG]')\n",
    "lims = [0, 50]\n",
    "plt.xlim(lims)\n",
    "plt.ylim(lims)\n",
    "_ = plt.plot(lims, lims)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "error1 = test_predictions1 - test_labels\n",
    "plt.hist(error1, bins = 25)\n",
    "plt.xlabel(\"Prediction Error [MPG]\")\n",
    "_ = plt.ylabel(\"Count\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "error2 = test_predictions2 - test_labels\n",
    "plt.hist(error2, bins = 25)\n",
    "plt.xlabel(\"Prediction Error [MPG]\")\n",
    "_ = plt.ylabel(\"Count\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It looks like we're off by a somewhat normal fashion in each of the error sets, but each prediction set is limited by its sample size. What if we were to combine the two error sets together?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "combined_errors = pd.concat([error1, error2])\n",
    "plt.hist(combined_errors, bins=25)\n",
    "plt.xlabel(\"Prediction Error [MPG]\")\n",
    "_ = plt.ylabel(\"Count\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Central Limit Theorem at work!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe src=\"https://en.wikipedia.org/wiki/Central_limit_theorem\" width=\"900\" height=\"800\"></iframe>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%HTML\n",
    "<iframe src=\"https://en.wikipedia.org/wiki/Central_limit_theorem\" width=\"900\" height=\"800\"></iframe>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Takeaways:\n",
    "- Neural networks are subject to _uncertainty_ in their weights due to the random nature of their initialization and the selection of training/validation data\n",
    "- This can lead to different weight selections in repeated applications of the same training routine.\n",
    "- This property is present, even when tricks like Early Stopping are used to control for overfitting."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Adding dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_dropout_model() -> tf.keras.Model:\n",
    "    model = keras.Sequential([\n",
    "        layers.Dense(64, activation='relu', input_shape=[len(train_dataset.keys())]),\n",
    "        layers.Dropout(0.5),\n",
    "        layers.Dense(64, activation='relu'),\n",
    "        layers.Dropout(0.5),\n",
    "        layers.Dense(1),\n",
    "    ])\n",
    "\n",
    "    optimizer = tf.keras.optimizers.RMSprop(0.001)\n",
    "\n",
    "    model.compile(loss='mse',\n",
    "                optimizer=optimizer,\n",
    "                metrics=['mae', 'mse'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "....................................................................................................\n",
      "...................................................................................................."
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dropout_model = build_dropout_model()\n",
    "\n",
    "history_dropout = dropout_model.fit(normed_train_data, train_labels, epochs=EPOCHS,\n",
    "                    validation_split=0.2, verbose=0, callbacks=[PrintDot()])\n",
    "\n",
    "plot_history(history_dropout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_dropout_predictions = dropout_model.predict(normed_test_data).flatten()\n",
    "\n",
    "dropout_error = test_dropout_predictions - test_labels\n",
    "plt.hist(dropout_error, bins = 25)\n",
    "plt.xlabel(\"Prediction Error [MPG]\")\n",
    "_ = plt.ylabel(\"Count\")"
   ]
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}