TF 1 Tutorials/overfitting_and_underfitting.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### [Explore overfitting and underfitting](https://www.tensorflow.org/tutorials/keras/overfit_and_underfit)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The most common ways to prevent overfitting in neural networks:\n",
    "- Get more training data,\n",
    "- **Reduce the capacity of the network**,\n",
    "- **Add weight regularization**,\n",
    "- **Add dropout**,\n",
    "- Data augmentation,\n",
    "- Batch normalization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import absolute_import, division, print_function, unicode_literals\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.12.0\n"
     ]
    }
   ],
   "source": [
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Load the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "NUM_WORDS = 10000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "(train_data, train_labels), (test_data, test_labels) = keras.datasets.imdb.load_data(num_words=NUM_WORDS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def multi_hot_seq(sequences, dimension):\n",
    "    results = np.zeros((len(sequences), dimension))\n",
    "    for i, word_indices in enumerate(sequences):\n",
    "        results[i, word_indices] = 1.0\n",
    "    return results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = multi_hot_seq(train_data, dimension=NUM_WORDS)\n",
    "test_data = multi_hot_seq(test_data, dimension=NUM_WORDS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1de97bb8d68>]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.plot(train_data[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Baseline model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "baseline_model = keras.Sequential([\n",
    "    keras.layers.Dense(16, activation=tf.nn.relu, input_shape=(NUM_WORDS,)),\n",
    "    keras.layers.Dense(16, activation=tf.nn.relu),\n",
    "    keras.layers.Dense(1, activation=tf.nn.sigmoid)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "baseline_model.compile(optimizer='adam',\n",
    "                      loss='binary_crossentropy',\n",
    "                      metrics=['accuracy', 'binary_crossentropy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 16)                160016    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 16)                272       \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 1)                 17        \n",
      "=================================================================\n",
      "Total params: 160,305\n",
      "Trainable params: 160,305\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "baseline_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 14s - loss: 0.4427 - acc: 0.8203 - binary_crossentropy: 0.4427 - val_loss: 0.3084 - val_acc: 0.8813 - val_binary_crossentropy: 0.3084\n",
      "Epoch 2/20\n",
      " - 13s - loss: 0.2251 - acc: 0.9169 - binary_crossentropy: 0.2251 - val_loss: 0.2862 - val_acc: 0.8860 - val_binary_crossentropy: 0.2862\n",
      "Epoch 3/20\n",
      " - 8s - loss: 0.1670 - acc: 0.9417 - binary_crossentropy: 0.1670 - val_loss: 0.3010 - val_acc: 0.8825 - val_binary_crossentropy: 0.3010\n",
      "Epoch 4/20\n",
      " - 8s - loss: 0.1335 - acc: 0.9544 - binary_crossentropy: 0.1335 - val_loss: 0.3334 - val_acc: 0.8767 - val_binary_crossentropy: 0.3334\n",
      "Epoch 5/20\n",
      " - 8s - loss: 0.1120 - acc: 0.9626 - binary_crossentropy: 0.1120 - val_loss: 0.3666 - val_acc: 0.8704 - val_binary_crossentropy: 0.3666\n",
      "Epoch 6/20\n",
      " - 9s - loss: 0.0924 - acc: 0.9720 - binary_crossentropy: 0.0924 - val_loss: 0.4013 - val_acc: 0.8663 - val_binary_crossentropy: 0.4013\n",
      "Epoch 7/20\n",
      " - 9s - loss: 0.0761 - acc: 0.9785 - binary_crossentropy: 0.0761 - val_loss: 0.4559 - val_acc: 0.8586 - val_binary_crossentropy: 0.4559\n",
      "Epoch 8/20\n",
      " - 10s - loss: 0.0639 - acc: 0.9823 - binary_crossentropy: 0.0639 - val_loss: 0.4889 - val_acc: 0.8598 - val_binary_crossentropy: 0.4889\n",
      "Epoch 9/20\n",
      " - 10s - loss: 0.0528 - acc: 0.9873 - binary_crossentropy: 0.0528 - val_loss: 0.5311 - val_acc: 0.8562 - val_binary_crossentropy: 0.5311\n",
      "Epoch 10/20\n",
      " - 11s - loss: 0.0438 - acc: 0.9908 - binary_crossentropy: 0.0438 - val_loss: 0.5832 - val_acc: 0.8525 - val_binary_crossentropy: 0.5832\n",
      "Epoch 11/20\n",
      " - 11s - loss: 0.0383 - acc: 0.9918 - binary_crossentropy: 0.0383 - val_loss: 0.6251 - val_acc: 0.8529 - val_binary_crossentropy: 0.6251\n",
      "Epoch 12/20\n",
      " - 9s - loss: 0.0297 - acc: 0.9946 - binary_crossentropy: 0.0297 - val_loss: 0.6706 - val_acc: 0.8514 - val_binary_crossentropy: 0.6706\n",
      "Epoch 13/20\n",
      " - 9s - loss: 0.0233 - acc: 0.9962 - binary_crossentropy: 0.0233 - val_loss: 0.7119 - val_acc: 0.8486 - val_binary_crossentropy: 0.7119\n",
      "Epoch 14/20\n",
      " - 9s - loss: 0.0185 - acc: 0.9978 - binary_crossentropy: 0.0185 - val_loss: 0.7527 - val_acc: 0.8477 - val_binary_crossentropy: 0.7527\n",
      "Epoch 15/20\n",
      " - 9s - loss: 0.0145 - acc: 0.9989 - binary_crossentropy: 0.0145 - val_loss: 0.7982 - val_acc: 0.8468 - val_binary_crossentropy: 0.7982\n",
      "Epoch 16/20\n",
      " - 10s - loss: 0.0112 - acc: 0.9992 - binary_crossentropy: 0.0112 - val_loss: 0.8394 - val_acc: 0.8476 - val_binary_crossentropy: 0.8394\n",
      "Epoch 17/20\n",
      " - 11s - loss: 0.0090 - acc: 0.9998 - binary_crossentropy: 0.0090 - val_loss: 0.8768 - val_acc: 0.8450 - val_binary_crossentropy: 0.8768\n",
      "Epoch 18/20\n",
      " - 11s - loss: 0.0069 - acc: 0.9998 - binary_crossentropy: 0.0069 - val_loss: 0.9121 - val_acc: 0.8452 - val_binary_crossentropy: 0.9121\n",
      "Epoch 19/20\n",
      " - 11s - loss: 0.0053 - acc: 0.9999 - binary_crossentropy: 0.0053 - val_loss: 0.9450 - val_acc: 0.8464 - val_binary_crossentropy: 0.9450\n",
      "Epoch 20/20\n",
      " - 10s - loss: 0.0043 - acc: 1.0000 - binary_crossentropy: 0.0043 - val_loss: 0.9780 - val_acc: 0.8452 - val_binary_crossentropy: 0.9780\n"
     ]
    }
   ],
   "source": [
    "baseline_history = baseline_model.fit(train_data, train_labels, epochs=20, batch_size=512, validation_data=(test_data, test_labels), verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Smaller model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "small_model = keras.Sequential([\n",
    "    keras.layers.Dense(4, activation=tf.nn.relu, input_shape=(NUM_WORDS,)),\n",
    "    keras.layers.Dense(4, activation=tf.nn.relu),\n",
    "    keras.layers.Dense(1, activation=tf.nn.sigmoid)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "small_model.compile(optimizer='adam',\n",
    "                   loss='binary_crossentropy',\n",
    "                   metrics=['accuracy', 'binary_crossentropy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_3 (Dense)              (None, 4)                 40004     \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              (None, 4)                 20        \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              (None, 1)                 5         \n",
      "=================================================================\n",
      "Total params: 40,029\n",
      "Trainable params: 40,029\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "small_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 16s - loss: 0.6163 - acc: 0.7622 - binary_crossentropy: 0.6163 - val_loss: 0.5252 - val_acc: 0.8348 - val_binary_crossentropy: 0.5252\n",
      "Epoch 2/20\n",
      " - 9s - loss: 0.4236 - acc: 0.8708 - binary_crossentropy: 0.4236 - val_loss: 0.3801 - val_acc: 0.8682 - val_binary_crossentropy: 0.3801\n",
      "Epoch 3/20\n",
      " - 9s - loss: 0.3038 - acc: 0.8996 - binary_crossentropy: 0.3038 - val_loss: 0.3168 - val_acc: 0.8828 - val_binary_crossentropy: 0.3168\n",
      "Epoch 4/20\n",
      " - 9s - loss: 0.2425 - acc: 0.9182 - binary_crossentropy: 0.2425 - val_loss: 0.2926 - val_acc: 0.8863 - val_binary_crossentropy: 0.2926\n",
      "Epoch 5/20\n",
      " - 9s - loss: 0.2060 - acc: 0.9304 - binary_crossentropy: 0.2060 - val_loss: 0.2851 - val_acc: 0.8872 - val_binary_crossentropy: 0.2851\n",
      "Epoch 6/20\n",
      " - 10s - loss: 0.1800 - acc: 0.9389 - binary_crossentropy: 0.1800 - val_loss: 0.2827 - val_acc: 0.8879 - val_binary_crossentropy: 0.2827\n",
      "Epoch 7/20\n",
      " - 9s - loss: 0.1601 - acc: 0.9469 - binary_crossentropy: 0.1601 - val_loss: 0.2870 - val_acc: 0.8867 - val_binary_crossentropy: 0.2870\n",
      "Epoch 8/20\n",
      " - 10s - loss: 0.1438 - acc: 0.9523 - binary_crossentropy: 0.1438 - val_loss: 0.2947 - val_acc: 0.8853 - val_binary_crossentropy: 0.2947\n",
      "Epoch 9/20\n",
      " - 10s - loss: 0.1309 - acc: 0.9567 - binary_crossentropy: 0.1309 - val_loss: 0.3045 - val_acc: 0.8828 - val_binary_crossentropy: 0.3045\n",
      "Epoch 10/20\n",
      " - 11s - loss: 0.1185 - acc: 0.9623 - binary_crossentropy: 0.1185 - val_loss: 0.3184 - val_acc: 0.8783 - val_binary_crossentropy: 0.3184\n",
      "Epoch 11/20\n",
      " - 12s - loss: 0.1083 - acc: 0.9667 - binary_crossentropy: 0.1083 - val_loss: 0.3297 - val_acc: 0.8778 - val_binary_crossentropy: 0.3297\n",
      "Epoch 12/20\n",
      " - 11s - loss: 0.0987 - acc: 0.9702 - binary_crossentropy: 0.0987 - val_loss: 0.3470 - val_acc: 0.8740 - val_binary_crossentropy: 0.3470\n",
      "Epoch 13/20\n",
      " - 9s - loss: 0.0901 - acc: 0.9739 - binary_crossentropy: 0.0901 - val_loss: 0.3595 - val_acc: 0.8731 - val_binary_crossentropy: 0.3595\n",
      "Epoch 14/20\n",
      " - 9s - loss: 0.0827 - acc: 0.9768 - binary_crossentropy: 0.0827 - val_loss: 0.3765 - val_acc: 0.8706 - val_binary_crossentropy: 0.3765\n",
      "Epoch 15/20\n",
      " - 9s - loss: 0.0755 - acc: 0.9797 - binary_crossentropy: 0.0755 - val_loss: 0.3932 - val_acc: 0.8698 - val_binary_crossentropy: 0.3932\n",
      "Epoch 16/20\n",
      " - 11s - loss: 0.0689 - acc: 0.9823 - binary_crossentropy: 0.0689 - val_loss: 0.4116 - val_acc: 0.8683 - val_binary_crossentropy: 0.4116\n",
      "Epoch 17/20\n",
      " - 9s - loss: 0.0630 - acc: 0.9848 - binary_crossentropy: 0.0630 - val_loss: 0.4308 - val_acc: 0.8658 - val_binary_crossentropy: 0.4308\n",
      "Epoch 18/20\n",
      " - 9s - loss: 0.0572 - acc: 0.9873 - binary_crossentropy: 0.0572 - val_loss: 0.4528 - val_acc: 0.8635 - val_binary_crossentropy: 0.4528\n",
      "Epoch 19/20\n",
      " - 9s - loss: 0.0522 - acc: 0.9892 - binary_crossentropy: 0.0522 - val_loss: 0.4692 - val_acc: 0.8638 - val_binary_crossentropy: 0.4692\n",
      "Epoch 20/20\n",
      " - 10s - loss: 0.0474 - acc: 0.9906 - binary_crossentropy: 0.0474 - val_loss: 0.4902 - val_acc: 0.8623 - val_binary_crossentropy: 0.4902\n"
     ]
    }
   ],
   "source": [
    "small_history = small_model.fit(train_data, train_labels, epochs=20, batch_size=512, validation_data=(test_data, test_labels), verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Larger model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "large_model = keras.Sequential([\n",
    "    keras.layers.Dense(512, activation=tf.nn.relu, input_shape=(NUM_WORDS,)),\n",
    "    keras.layers.Dense(512, activation=tf.nn.relu),\n",
    "    keras.layers.Dense(1, activation=tf.nn.sigmoid)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "large_model.compile(optimizer='adam',\n",
    "                   loss='binary_crossentropy',\n",
    "                   metrics=['accuracy', 'binary_crossentropy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_6 (Dense)              (None, 512)               5120512   \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 512)               262656    \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              (None, 1)                 513       \n",
      "=================================================================\n",
      "Total params: 5,383,681\n",
      "Trainable params: 5,383,681\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "large_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 52s - loss: 0.3481 - acc: 0.8524 - binary_crossentropy: 0.3481 - val_loss: 0.2969 - val_acc: 0.8801 - val_binary_crossentropy: 0.2969\n",
      "Epoch 2/20\n",
      " - 49s - loss: 0.1413 - acc: 0.9488 - binary_crossentropy: 0.1413 - val_loss: 0.3565 - val_acc: 0.8645 - val_binary_crossentropy: 0.3565\n",
      "Epoch 3/20\n",
      " - 49s - loss: 0.0493 - acc: 0.9856 - binary_crossentropy: 0.0493 - val_loss: 0.4309 - val_acc: 0.8676 - val_binary_crossentropy: 0.4309\n",
      "Epoch 4/20\n",
      " - 49s - loss: 0.0081 - acc: 0.9990 - binary_crossentropy: 0.0081 - val_loss: 0.5634 - val_acc: 0.8709 - val_binary_crossentropy: 0.5634\n",
      "Epoch 5/20\n",
      " - 49s - loss: 9.2941e-04 - acc: 1.0000 - binary_crossentropy: 9.2941e-04 - val_loss: 0.6577 - val_acc: 0.8713 - val_binary_crossentropy: 0.6577\n",
      "Epoch 6/20\n",
      " - 55s - loss: 2.4597e-04 - acc: 1.0000 - binary_crossentropy: 2.4597e-04 - val_loss: 0.7015 - val_acc: 0.8706 - val_binary_crossentropy: 0.7015\n",
      "Epoch 7/20\n",
      " - 52s - loss: 1.4286e-04 - acc: 1.0000 - binary_crossentropy: 1.4286e-04 - val_loss: 0.7309 - val_acc: 0.8708 - val_binary_crossentropy: 0.7309\n",
      "Epoch 8/20\n",
      " - 54s - loss: 9.5949e-05 - acc: 1.0000 - binary_crossentropy: 9.5949e-05 - val_loss: 0.7564 - val_acc: 0.8712 - val_binary_crossentropy: 0.7564\n",
      "Epoch 9/20\n",
      " - 53s - loss: 6.7637e-05 - acc: 1.0000 - binary_crossentropy: 6.7637e-05 - val_loss: 0.7804 - val_acc: 0.8714 - val_binary_crossentropy: 0.7804\n",
      "Epoch 10/20\n",
      " - 52s - loss: 4.9397e-05 - acc: 1.0000 - binary_crossentropy: 4.9397e-05 - val_loss: 0.8011 - val_acc: 0.8715 - val_binary_crossentropy: 0.8011\n",
      "Epoch 11/20\n",
      " - 55s - loss: 3.7220e-05 - acc: 1.0000 - binary_crossentropy: 3.7220e-05 - val_loss: 0.8205 - val_acc: 0.8715 - val_binary_crossentropy: 0.8205\n",
      "Epoch 12/20\n",
      " - 52s - loss: 2.8740e-05 - acc: 1.0000 - binary_crossentropy: 2.8740e-05 - val_loss: 0.8382 - val_acc: 0.8714 - val_binary_crossentropy: 0.8382\n",
      "Epoch 13/20\n",
      " - 54s - loss: 2.2755e-05 - acc: 1.0000 - binary_crossentropy: 2.2755e-05 - val_loss: 0.8532 - val_acc: 0.8714 - val_binary_crossentropy: 0.8532\n",
      "Epoch 14/20\n",
      " - 54s - loss: 1.8401e-05 - acc: 1.0000 - binary_crossentropy: 1.8401e-05 - val_loss: 0.8680 - val_acc: 0.8713 - val_binary_crossentropy: 0.8680\n",
      "Epoch 15/20\n",
      " - 51s - loss: 1.5133e-05 - acc: 1.0000 - binary_crossentropy: 1.5133e-05 - val_loss: 0.8807 - val_acc: 0.8712 - val_binary_crossentropy: 0.8807\n",
      "Epoch 16/20\n",
      " - 53s - loss: 1.2653e-05 - acc: 1.0000 - binary_crossentropy: 1.2653e-05 - val_loss: 0.8921 - val_acc: 0.8713 - val_binary_crossentropy: 0.8921\n",
      "Epoch 17/20\n",
      " - 54s - loss: 1.0699e-05 - acc: 1.0000 - binary_crossentropy: 1.0699e-05 - val_loss: 0.9041 - val_acc: 0.8713 - val_binary_crossentropy: 0.9041\n",
      "Epoch 18/20\n",
      " - 55s - loss: 9.1793e-06 - acc: 1.0000 - binary_crossentropy: 9.1793e-06 - val_loss: 0.9134 - val_acc: 0.8716 - val_binary_crossentropy: 0.9134\n",
      "Epoch 19/20\n",
      " - 53s - loss: 7.9268e-06 - acc: 1.0000 - binary_crossentropy: 7.9268e-06 - val_loss: 0.9226 - val_acc: 0.8715 - val_binary_crossentropy: 0.9226\n",
      "Epoch 20/20\n",
      " - 53s - loss: 6.9263e-06 - acc: 1.0000 - binary_crossentropy: 6.9263e-06 - val_loss: 0.9317 - val_acc: 0.8712 - val_binary_crossentropy: 0.9317\n"
     ]
    }
   ],
   "source": [
    "large_history = large_model.fit(train_data, train_labels, epochs=20, batch_size=512, validation_data=(test_data, test_labels), verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Plot histories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot(histories, key='binary_crossentropy'):\n",
    "    plt.figure(figsize=(16, 10))\n",
    "    \n",
    "    for name, history in histories:\n",
    "        val = plt.plot(history.epoch, history.history['val_'+key], ':', label=name.title()+' Validation')\n",
    "        plt.plot(history.epoch, history.history[key], label=name.title()+' Train', color=val[0].get_color())\n",
    "        \n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel(key.replace('_', ' ').title())\n",
    "    plt.xlim(0, max(history.epoch))\n",
    "    plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot([('baseline', baseline_history), ('smaller', small_history), ('larger', large_history)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Weigh regularization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "l2_model = keras.Sequential([\n",
    "    keras.layers.Dense(16, kernel_regularizer=keras.regularizers.l2(0.001), activation=tf.nn.relu, input_shape=(NUM_WORDS,)),\n",
    "    keras.layers.Dense(16, kernel_regularizer=keras.regularizers.l2(0.001), activation=tf.nn.relu),\n",
    "    keras.layers.Dense(1, activation=tf.nn.sigmoid)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "l2_model.compile(optimizer='adam',\n",
    "                loss='binary_crossentropy',\n",
    "                metrics=['accuracy', 'binary_crossentropy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 28s - loss: 0.5016 - acc: 0.8120 - binary_crossentropy: 0.4593 - val_loss: 0.3687 - val_acc: 0.8796 - val_binary_crossentropy: 0.3232\n",
      "Epoch 2/20\n",
      " - 12s - loss: 0.2971 - acc: 0.9130 - binary_crossentropy: 0.2476 - val_loss: 0.3362 - val_acc: 0.8875 - val_binary_crossentropy: 0.2841\n",
      "Epoch 3/20\n",
      " - 12s - loss: 0.2527 - acc: 0.9314 - binary_crossentropy: 0.1986 - val_loss: 0.3406 - val_acc: 0.8859 - val_binary_crossentropy: 0.2850\n",
      "Epoch 4/20\n",
      " - 10s - loss: 0.2298 - acc: 0.9406 - binary_crossentropy: 0.1727 - val_loss: 0.3541 - val_acc: 0.8819 - val_binary_crossentropy: 0.2962\n",
      "Epoch 5/20\n",
      " - 10s - loss: 0.2181 - acc: 0.9461 - binary_crossentropy: 0.1587 - val_loss: 0.3660 - val_acc: 0.8790 - val_binary_crossentropy: 0.3057\n",
      "Epoch 6/20\n",
      " - 10s - loss: 0.2059 - acc: 0.9508 - binary_crossentropy: 0.1451 - val_loss: 0.3821 - val_acc: 0.8750 - val_binary_crossentropy: 0.3210\n",
      "Epoch 7/20\n",
      " - 10s - loss: 0.1995 - acc: 0.9537 - binary_crossentropy: 0.1373 - val_loss: 0.3952 - val_acc: 0.8729 - val_binary_crossentropy: 0.3324\n",
      "Epoch 8/20\n",
      " - 11s - loss: 0.1927 - acc: 0.9574 - binary_crossentropy: 0.1293 - val_loss: 0.4078 - val_acc: 0.8706 - val_binary_crossentropy: 0.3441\n",
      "Epoch 9/20\n",
      " - 11s - loss: 0.1881 - acc: 0.9568 - binary_crossentropy: 0.1239 - val_loss: 0.4251 - val_acc: 0.8697 - val_binary_crossentropy: 0.3604\n",
      "Epoch 10/20\n",
      " - 12s - loss: 0.1866 - acc: 0.9583 - binary_crossentropy: 0.1212 - val_loss: 0.4373 - val_acc: 0.8656 - val_binary_crossentropy: 0.3715\n",
      "Epoch 11/20\n",
      " - 11s - loss: 0.1823 - acc: 0.9606 - binary_crossentropy: 0.1161 - val_loss: 0.4487 - val_acc: 0.8654 - val_binary_crossentropy: 0.3819\n",
      "Epoch 12/20\n",
      " - 12s - loss: 0.1792 - acc: 0.9614 - binary_crossentropy: 0.1124 - val_loss: 0.4710 - val_acc: 0.8599 - val_binary_crossentropy: 0.4040\n",
      "Epoch 13/20\n",
      " - 10s - loss: 0.1789 - acc: 0.9606 - binary_crossentropy: 0.1110 - val_loss: 0.4693 - val_acc: 0.8622 - val_binary_crossentropy: 0.4011\n",
      "Epoch 14/20\n",
      " - 12s - loss: 0.1710 - acc: 0.9640 - binary_crossentropy: 0.1029 - val_loss: 0.4784 - val_acc: 0.8594 - val_binary_crossentropy: 0.4108\n",
      "Epoch 15/20\n",
      " - 14s - loss: 0.1653 - acc: 0.9676 - binary_crossentropy: 0.0977 - val_loss: 0.4917 - val_acc: 0.8590 - val_binary_crossentropy: 0.4241\n",
      "Epoch 16/20\n",
      " - 13s - loss: 0.1642 - acc: 0.9677 - binary_crossentropy: 0.0962 - val_loss: 0.5036 - val_acc: 0.8572 - val_binary_crossentropy: 0.4354\n",
      "Epoch 17/20\n",
      " - 12s - loss: 0.1628 - acc: 0.9678 - binary_crossentropy: 0.0946 - val_loss: 0.5145 - val_acc: 0.8553 - val_binary_crossentropy: 0.4459\n",
      "Epoch 18/20\n",
      " - 12s - loss: 0.1624 - acc: 0.9676 - binary_crossentropy: 0.0935 - val_loss: 0.5332 - val_acc: 0.8543 - val_binary_crossentropy: 0.4639\n",
      "Epoch 19/20\n",
      " - 13s - loss: 0.1606 - acc: 0.9676 - binary_crossentropy: 0.0912 - val_loss: 0.5311 - val_acc: 0.8583 - val_binary_crossentropy: 0.4615\n",
      "Epoch 20/20\n",
      " - 12s - loss: 0.1601 - acc: 0.9682 - binary_crossentropy: 0.0900 - val_loss: 0.5450 - val_acc: 0.8544 - val_binary_crossentropy: 0.4746\n"
     ]
    }
   ],
   "source": [
    "l2_model_history = l2_model.fit(train_data, train_labels, epochs=20, batch_size=512, validation_data=(test_data, test_labels), verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot([('baseline', baseline_history), ('l2', l2_model_history)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "dpt_model = keras.Sequential([\n",
    "    keras.layers.Dense(16, activation=tf.nn.relu, input_shape=(NUM_WORDS,)),\n",
    "    keras.layers.Dropout(0.5),\n",
    "    keras.layers.Dense(16, activation=tf.nn.relu),\n",
    "    keras.layers.Dropout(0.5),\n",
    "    keras.layers.Dense(1, activation=tf.nn.sigmoid)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "dpt_model.compile(optimizer='adam',\n",
    "                 loss='binary_crossentropy',\n",
    "                 metrics=['accuracy', 'binary_crossentropy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 27s - loss: 0.6366 - acc: 0.6220 - binary_crossentropy: 0.6366 - val_loss: 0.4959 - val_acc: 0.8526 - val_binary_crossentropy: 0.4959\n",
      "Epoch 2/20\n",
      " - 11s - loss: 0.4686 - acc: 0.7984 - binary_crossentropy: 0.4686 - val_loss: 0.3408 - val_acc: 0.8783 - val_binary_crossentropy: 0.3408\n",
      "Epoch 3/20\n",
      " - 13s - loss: 0.3620 - acc: 0.8680 - binary_crossentropy: 0.3620 - val_loss: 0.2894 - val_acc: 0.8873 - val_binary_crossentropy: 0.2894\n",
      "Epoch 4/20\n",
      " - 14s - loss: 0.2892 - acc: 0.8987 - binary_crossentropy: 0.2892 - val_loss: 0.2739 - val_acc: 0.8898 - val_binary_crossentropy: 0.2739\n",
      "Epoch 5/20\n",
      " - 13s - loss: 0.2490 - acc: 0.9169 - binary_crossentropy: 0.2490 - val_loss: 0.2740 - val_acc: 0.8876 - val_binary_crossentropy: 0.2740\n",
      "Epoch 6/20\n",
      " - 12s - loss: 0.2200 - acc: 0.9263 - binary_crossentropy: 0.2200 - val_loss: 0.2859 - val_acc: 0.8861 - val_binary_crossentropy: 0.2859\n",
      "Epoch 7/20\n",
      " - 13s - loss: 0.1909 - acc: 0.9381 - binary_crossentropy: 0.1909 - val_loss: 0.3062 - val_acc: 0.8862 - val_binary_crossentropy: 0.3062\n",
      "Epoch 8/20\n",
      " - 15s - loss: 0.1714 - acc: 0.9437 - binary_crossentropy: 0.1714 - val_loss: 0.3134 - val_acc: 0.8842 - val_binary_crossentropy: 0.3134\n",
      "Epoch 9/20\n",
      " - 12s - loss: 0.1540 - acc: 0.9493 - binary_crossentropy: 0.1540 - val_loss: 0.3438 - val_acc: 0.8821 - val_binary_crossentropy: 0.3438\n",
      "Epoch 10/20\n",
      " - 12s - loss: 0.1384 - acc: 0.9540 - binary_crossentropy: 0.1384 - val_loss: 0.3570 - val_acc: 0.8807 - val_binary_crossentropy: 0.3570\n",
      "Epoch 11/20\n",
      " - 16s - loss: 0.1282 - acc: 0.9557 - binary_crossentropy: 0.1282 - val_loss: 0.3667 - val_acc: 0.8798 - val_binary_crossentropy: 0.3667\n",
      "Epoch 12/20\n",
      " - 9s - loss: 0.1139 - acc: 0.9622 - binary_crossentropy: 0.1139 - val_loss: 0.4047 - val_acc: 0.8788 - val_binary_crossentropy: 0.4047\n",
      "Epoch 13/20\n",
      " - 9s - loss: 0.1064 - acc: 0.9641 - binary_crossentropy: 0.1064 - val_loss: 0.4432 - val_acc: 0.8767 - val_binary_crossentropy: 0.4432\n",
      "Epoch 14/20\n",
      "WARNING:tensorflow:Method on_batch_end() is slow compared to the batch update (0.740787). Check your callbacks.\n",
      "WARNING:tensorflow:Method on_batch_end() is slow compared to the batch update (0.370394). Check your callbacks.\n",
      " - 23s - loss: 0.0987 - acc: 0.9660 - binary_crossentropy: 0.0987 - val_loss: 0.4215 - val_acc: 0.8778 - val_binary_crossentropy: 0.4215\n",
      "Epoch 15/20\n",
      " - 10s - loss: 0.0969 - acc: 0.9653 - binary_crossentropy: 0.0969 - val_loss: 0.4473 - val_acc: 0.8779 - val_binary_crossentropy: 0.4473\n",
      "Epoch 16/20\n",
      " - 11s - loss: 0.0878 - acc: 0.9696 - binary_crossentropy: 0.0878 - val_loss: 0.4786 - val_acc: 0.8762 - val_binary_crossentropy: 0.4786\n",
      "Epoch 17/20\n",
      " - 11s - loss: 0.0823 - acc: 0.9712 - binary_crossentropy: 0.0823 - val_loss: 0.4957 - val_acc: 0.8763 - val_binary_crossentropy: 0.4957\n",
      "Epoch 18/20\n",
      " - 10s - loss: 0.0798 - acc: 0.9709 - binary_crossentropy: 0.0798 - val_loss: 0.5050 - val_acc: 0.8758 - val_binary_crossentropy: 0.5050\n",
      "Epoch 19/20\n",
      " - 10s - loss: 0.0734 - acc: 0.9745 - binary_crossentropy: 0.0734 - val_loss: 0.5552 - val_acc: 0.8762 - val_binary_crossentropy: 0.5552\n",
      "Epoch 20/20\n",
      " - 10s - loss: 0.0729 - acc: 0.9722 - binary_crossentropy: 0.0729 - val_loss: 0.5562 - val_acc: 0.8747 - val_binary_crossentropy: 0.5562\n"
     ]
    }
   ],
   "source": [
    "dpt_model_history = dpt_model.fit(train_data, train_labels, epochs=20, batch_size=512, validation_data=(test_data, test_labels), verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot([('baseline', baseline_history), ('dropout', dpt_model_history)])"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}