NN with TensorFlow.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"The accuracy of our model on the validation data would peak after training for a number of epochs, and would then start decreasing. In other words, our model would overfit to the training data. Learning how to deal with overfitting is important. Although it's often possible to achieve high accuracy on the training set, what we really want is to develop models that generalize well to a testing set (or data they haven't seen before).\\n\\nThe opposite of overfitting is underfitting. Underfitting occurs when there is still room for improvement on the test data. This can happen for a number of reasons: If the model is not powerful enough, is over-regularized, or has simply not been trained long enough. This means the network has not learned the relevant patterns in the training data.\\n\\nIf you train for too long though, the model will start to overfit and learn patterns from the training data that don't generalize to the test data. We need to strike a balance. Understanding how to train for an appropriate number of epochs as we'll explore below is a useful skill.\\n\\nTo prevent overfitting, the best solution is to use more training data. A model trained on more data will naturally generalize better. When that is no longer possible, the next best solution is to use techniques like regularization. These place constraints on the quantity and type of information your model can store. If a network can only afford to memorize a small number of patterns, the optimization process will force it to focus on the most prominent patterns, which have a better chance of generalizing well.\\n\\nIn this notebook, we'll explore two common regularization techniques—weight regularization and dropout—and use them to improve our IMDB movie review classification notebook.\\nRather than using an embedding as in the previous notebook, here we will multi-hot encode the sentences. This model will quickly overfit to the training set. It will be used to demonstrate when overfitting occurs, and how to fight it.\\n\\nMulti-hot-encoding our lists means turning them into vectors of 0s and 1s. Concretely, this would mean for instance turning the sequence [3, 5] into a 10,000-dimensional vector that would be all-zeros except for indices 3 and 5, which would be ones.\""
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'''The accuracy of our model on the validation data would peak after training for a number of epochs, and would then start decreasing. In other words, our model would overfit to the training data. Learning how to deal with overfitting is important. Although it's often possible to achieve high accuracy on the training set, what we really want is to develop models that generalize well to a testing set (or data they haven't seen before).\n",
    "\n",
    "The opposite of overfitting is underfitting. Underfitting occurs when there is still room for improvement on the test data. This can happen for a number of reasons: If the model is not powerful enough, is over-regularized, or has simply not been trained long enough. This means the network has not learned the relevant patterns in the training data.\n",
    "\n",
    "If you train for too long though, the model will start to overfit and learn patterns from the training data that don't generalize to the test data. We need to strike a balance. Understanding how to train for an appropriate number of epochs as we'll explore below is a useful skill.\n",
    "\n",
    "To prevent overfitting, the best solution is to use more training data. A model trained on more data will naturally generalize better. When that is no longer possible, the next best solution is to use techniques like regularization. These place constraints on the quantity and type of information your model can store. If a network can only afford to memorize a small number of patterns, the optimization process will force it to focus on the most prominent patterns, which have a better chance of generalizing well.\n",
    "\n",
    "In this notebook, we'll explore two common regularization techniques—weight regularization and dropout—and use them to improve our IMDB movie review classification notebook.\n",
    "Rather than using an embedding as in the previous notebook, here we will multi-hot encode the sentences. This model will quickly overfit to the training set. It will be used to demonstrate when overfitting occurs, and how to fight it.\n",
    "\n",
    "Multi-hot-encoding our lists means turning them into vectors of 0s and 1s. Concretely, this would mean for instance turning the sequence [3, 5] into a 10,000-dimensional vector that would be all-zeros except for indices 3 and 5, which would be ones.'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.12.0\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "NUM_WORDS = 10000\n",
    "\n",
    "(train_data, train_labels), (test_data, test_labels) = keras.datasets.imdb.load_data(num_words=NUM_WORDS)\n",
    "\n",
    "def multi_hot_sequences(sequences, dimension):\n",
    "    # Create an all-zero matrix of shape (len(sequences), dimension)\n",
    "    results = np.zeros((len(sequences), dimension))\n",
    "    for i, word_indices in enumerate(sequences):\n",
    "        results[i, word_indices] = 1.0  # set specific indices of results[i] to 1s\n",
    "    return results\n",
    "\n",
    "\n",
    "train_data = multi_hot_sequences(train_data, dimension=NUM_WORDS)\n",
    "test_data = multi_hot_sequences(test_data, dimension=NUM_WORDS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x21682b8bc88>]"
      ]
     },
     "execution_count": 4,
     "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": "code",
   "execution_count": 5,
   "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 = keras.Sequential([\n",
    "    # `input_shape` is only required here so that `.summary` works.\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",
    "])\n",
    "\n",
    "baseline_model.compile(optimizer='adam',\n",
    "                       loss='binary_crossentropy',\n",
    "                       metrics=['accuracy', 'binary_crossentropy'])\n",
    "\n",
    "baseline_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 6s - loss: 0.5163 - acc: 0.8033 - binary_crossentropy: 0.5163 - val_loss: 0.3630 - val_acc: 0.8687 - val_binary_crossentropy: 0.3630\n",
      "Epoch 2/20\n",
      " - 5s - loss: 0.2677 - acc: 0.9054 - binary_crossentropy: 0.2677 - val_loss: 0.2860 - val_acc: 0.8876 - val_binary_crossentropy: 0.2860\n",
      "Epoch 3/20\n",
      " - 5s - loss: 0.1927 - acc: 0.9318 - binary_crossentropy: 0.1927 - val_loss: 0.2848 - val_acc: 0.8861 - val_binary_crossentropy: 0.2848\n",
      "Epoch 4/20\n",
      " - 5s - loss: 0.1532 - acc: 0.9478 - binary_crossentropy: 0.1532 - val_loss: 0.3059 - val_acc: 0.8804 - val_binary_crossentropy: 0.3059\n",
      "Epoch 5/20\n",
      " - 5s - loss: 0.1265 - acc: 0.9574 - binary_crossentropy: 0.1265 - val_loss: 0.3327 - val_acc: 0.8761 - val_binary_crossentropy: 0.3327\n",
      "Epoch 6/20\n",
      " - 5s - loss: 0.1054 - acc: 0.9665 - binary_crossentropy: 0.1054 - val_loss: 0.3621 - val_acc: 0.8720 - val_binary_crossentropy: 0.3621\n",
      "Epoch 7/20\n",
      " - 5s - loss: 0.0883 - acc: 0.9730 - binary_crossentropy: 0.0883 - val_loss: 0.4151 - val_acc: 0.8642 - val_binary_crossentropy: 0.4151\n",
      "Epoch 8/20\n",
      " - 5s - loss: 0.0743 - acc: 0.9779 - binary_crossentropy: 0.0743 - val_loss: 0.4411 - val_acc: 0.8649 - val_binary_crossentropy: 0.4411\n",
      "Epoch 9/20\n",
      " - 5s - loss: 0.0612 - acc: 0.9842 - binary_crossentropy: 0.0612 - val_loss: 0.4840 - val_acc: 0.8614 - val_binary_crossentropy: 0.4840\n",
      "Epoch 10/20\n",
      " - 5s - loss: 0.0502 - acc: 0.9888 - binary_crossentropy: 0.0502 - val_loss: 0.5343 - val_acc: 0.8565 - val_binary_crossentropy: 0.5343\n",
      "Epoch 11/20\n",
      " - 5s - loss: 0.0417 - acc: 0.9908 - binary_crossentropy: 0.0417 - val_loss: 0.5787 - val_acc: 0.8555 - val_binary_crossentropy: 0.5787\n",
      "Epoch 12/20\n",
      " - 5s - loss: 0.0329 - acc: 0.9938 - binary_crossentropy: 0.0329 - val_loss: 0.6276 - val_acc: 0.8550 - val_binary_crossentropy: 0.6276\n",
      "Epoch 13/20\n",
      " - 5s - loss: 0.0261 - acc: 0.9960 - binary_crossentropy: 0.0261 - val_loss: 0.6743 - val_acc: 0.8526 - val_binary_crossentropy: 0.6743\n",
      "Epoch 14/20\n",
      " - 5s - loss: 0.0206 - acc: 0.9973 - binary_crossentropy: 0.0206 - val_loss: 0.7200 - val_acc: 0.8506 - val_binary_crossentropy: 0.7200\n",
      "Epoch 15/20\n",
      " - 5s - loss: 0.0159 - acc: 0.9986 - binary_crossentropy: 0.0159 - val_loss: 0.7678 - val_acc: 0.8499 - val_binary_crossentropy: 0.7678\n",
      "Epoch 16/20\n",
      " - 5s - loss: 0.0123 - acc: 0.9991 - binary_crossentropy: 0.0123 - val_loss: 0.8088 - val_acc: 0.8507 - val_binary_crossentropy: 0.8088\n",
      "Epoch 17/20\n",
      " - 5s - loss: 0.0095 - acc: 0.9996 - binary_crossentropy: 0.0095 - val_loss: 0.8516 - val_acc: 0.8490 - val_binary_crossentropy: 0.8516\n",
      "Epoch 18/20\n",
      " - 5s - loss: 0.0074 - acc: 0.9998 - binary_crossentropy: 0.0074 - val_loss: 0.8873 - val_acc: 0.8486 - val_binary_crossentropy: 0.8873\n",
      "Epoch 19/20\n",
      " - 5s - loss: 0.0058 - acc: 0.9999 - binary_crossentropy: 0.0058 - val_loss: 0.9178 - val_acc: 0.8492 - val_binary_crossentropy: 0.9178\n",
      "Epoch 20/20\n",
      " - 5s - loss: 0.0047 - acc: 1.0000 - binary_crossentropy: 0.0047 - val_loss: 0.9494 - val_acc: 0.8482 - val_binary_crossentropy: 0.9494\n"
     ]
    }
   ],
   "source": [
    "baseline_history = baseline_model.fit(train_data,\n",
    "                                      train_labels,\n",
    "                                      epochs=20,\n",
    "                                      batch_size=512,\n",
    "                                      validation_data=(test_data, test_labels),\n",
    "                                      verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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": [
    "smaller_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",
    "])\n",
    "\n",
    "smaller_model.compile(optimizer='adam',\n",
    "                loss='binary_crossentropy',\n",
    "                metrics=['accuracy', 'binary_crossentropy'])\n",
    "\n",
    "smaller_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 5s - loss: 0.6540 - acc: 0.6568 - binary_crossentropy: 0.6540 - val_loss: 0.6134 - val_acc: 0.7310 - val_binary_crossentropy: 0.6134\n",
      "Epoch 2/20\n",
      " - 5s - loss: 0.5694 - acc: 0.7774 - binary_crossentropy: 0.5694 - val_loss: 0.5537 - val_acc: 0.7741 - val_binary_crossentropy: 0.5537\n",
      "Epoch 3/20\n",
      " - 5s - loss: 0.5104 - acc: 0.8342 - binary_crossentropy: 0.5104 - val_loss: 0.5119 - val_acc: 0.8356 - val_binary_crossentropy: 0.5119\n",
      "Epoch 4/20\n",
      " - 5s - loss: 0.4674 - acc: 0.8692 - binary_crossentropy: 0.4674 - val_loss: 0.4823 - val_acc: 0.8536 - val_binary_crossentropy: 0.4823\n",
      "Epoch 5/20\n",
      " - 5s - loss: 0.4341 - acc: 0.8906 - binary_crossentropy: 0.4341 - val_loss: 0.4607 - val_acc: 0.8604 - val_binary_crossentropy: 0.4607\n",
      "Epoch 6/20\n",
      " - 5s - loss: 0.4065 - acc: 0.9065 - binary_crossentropy: 0.4065 - val_loss: 0.4448 - val_acc: 0.8644 - val_binary_crossentropy: 0.4448\n",
      "Epoch 7/20\n",
      " - 6s - loss: 0.3832 - acc: 0.9177 - binary_crossentropy: 0.3832 - val_loss: 0.4312 - val_acc: 0.8716 - val_binary_crossentropy: 0.4312\n",
      "Epoch 8/20\n",
      " - 6s - loss: 0.3623 - acc: 0.9274 - binary_crossentropy: 0.3623 - val_loss: 0.4204 - val_acc: 0.8752 - val_binary_crossentropy: 0.4204\n",
      "Epoch 9/20\n",
      " - 6s - loss: 0.3441 - acc: 0.9357 - binary_crossentropy: 0.3441 - val_loss: 0.4131 - val_acc: 0.8739 - val_binary_crossentropy: 0.4131\n",
      "Epoch 10/20\n",
      " - 6s - loss: 0.3271 - acc: 0.9416 - binary_crossentropy: 0.3271 - val_loss: 0.4036 - val_acc: 0.8817 - val_binary_crossentropy: 0.4036\n",
      "Epoch 11/20\n",
      " - 6s - loss: 0.3117 - acc: 0.9467 - binary_crossentropy: 0.3117 - val_loss: 0.4005 - val_acc: 0.8782 - val_binary_crossentropy: 0.4005\n",
      "Epoch 12/20\n",
      " - 6s - loss: 0.2975 - acc: 0.9516 - binary_crossentropy: 0.2975 - val_loss: 0.3931 - val_acc: 0.8815 - val_binary_crossentropy: 0.3931\n",
      "Epoch 13/20\n",
      " - 6s - loss: 0.2843 - acc: 0.9558 - binary_crossentropy: 0.2843 - val_loss: 0.3944 - val_acc: 0.8769 - val_binary_crossentropy: 0.3944\n",
      "Epoch 14/20\n",
      " - 6s - loss: 0.2723 - acc: 0.9585 - binary_crossentropy: 0.2723 - val_loss: 0.3922 - val_acc: 0.8768 - val_binary_crossentropy: 0.3922\n",
      "Epoch 15/20\n",
      " - 6s - loss: 0.2603 - acc: 0.9624 - binary_crossentropy: 0.2603 - val_loss: 0.3869 - val_acc: 0.8775 - val_binary_crossentropy: 0.3869\n",
      "Epoch 16/20\n",
      " - 6s - loss: 0.2493 - acc: 0.9656 - binary_crossentropy: 0.2493 - val_loss: 0.3904 - val_acc: 0.8743 - val_binary_crossentropy: 0.3904\n",
      "Epoch 17/20\n",
      " - 6s - loss: 0.2394 - acc: 0.9685 - binary_crossentropy: 0.2394 - val_loss: 0.3832 - val_acc: 0.8770 - val_binary_crossentropy: 0.3832\n",
      "Epoch 18/20\n",
      " - 7s - loss: 0.2296 - acc: 0.9708 - binary_crossentropy: 0.2296 - val_loss: 0.3848 - val_acc: 0.8752 - val_binary_crossentropy: 0.3848\n",
      "Epoch 19/20\n",
      " - 7s - loss: 0.2206 - acc: 0.9730 - binary_crossentropy: 0.2206 - val_loss: 0.3862 - val_acc: 0.8742 - val_binary_crossentropy: 0.3862\n",
      "Epoch 20/20\n",
      " - 6s - loss: 0.2119 - acc: 0.9742 - binary_crossentropy: 0.2119 - val_loss: 0.3946 - val_acc: 0.8719 - val_binary_crossentropy: 0.3946\n"
     ]
    }
   ],
   "source": [
    "smaller_history = smaller_model.fit(train_data,\n",
    "                                    train_labels,\n",
    "                                    epochs=20,\n",
    "                                    batch_size=512,\n",
    "                                    validation_data=(test_data, test_labels),\n",
    "                                    verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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": [
    "bigger_model = keras.models.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",
    "])\n",
    "\n",
    "bigger_model.compile(optimizer='adam',\n",
    "                     loss='binary_crossentropy',\n",
    "                     metrics=['accuracy','binary_crossentropy'])\n",
    "\n",
    "bigger_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 12s - loss: 0.3511 - acc: 0.8481 - binary_crossentropy: 0.3511 - val_loss: 0.3047 - val_acc: 0.8755 - val_binary_crossentropy: 0.3047\n",
      "Epoch 2/20\n",
      " - 11s - loss: 0.1417 - acc: 0.9485 - binary_crossentropy: 0.1417 - val_loss: 0.3576 - val_acc: 0.8640 - val_binary_crossentropy: 0.3576\n",
      "Epoch 3/20\n",
      " - 11s - loss: 0.0494 - acc: 0.9858 - binary_crossentropy: 0.0494 - val_loss: 0.4236 - val_acc: 0.8678 - val_binary_crossentropy: 0.4236\n",
      "Epoch 4/20\n",
      " - 12s - loss: 0.0080 - acc: 0.9988 - binary_crossentropy: 0.0080 - val_loss: 0.5729 - val_acc: 0.8707 - val_binary_crossentropy: 0.5729\n",
      "Epoch 5/20\n",
      " - 11s - loss: 0.0012 - acc: 1.0000 - binary_crossentropy: 0.0012 - val_loss: 0.6656 - val_acc: 0.8713 - val_binary_crossentropy: 0.6656\n",
      "Epoch 6/20\n",
      " - 11s - loss: 2.6661e-04 - acc: 1.0000 - binary_crossentropy: 2.6661e-04 - val_loss: 0.7117 - val_acc: 0.8711 - val_binary_crossentropy: 0.7117\n",
      "Epoch 7/20\n",
      " - 11s - loss: 1.5333e-04 - acc: 1.0000 - binary_crossentropy: 1.5333e-04 - val_loss: 0.7391 - val_acc: 0.8716 - val_binary_crossentropy: 0.7391\n",
      "Epoch 8/20\n",
      " - 11s - loss: 1.0686e-04 - acc: 1.0000 - binary_crossentropy: 1.0686e-04 - val_loss: 0.7605 - val_acc: 0.8718 - val_binary_crossentropy: 0.7605\n",
      "Epoch 9/20\n",
      " - 12s - loss: 7.9299e-05 - acc: 1.0000 - binary_crossentropy: 7.9299e-05 - val_loss: 0.7793 - val_acc: 0.8713 - val_binary_crossentropy: 0.7793\n",
      "Epoch 10/20\n",
      " - 11s - loss: 6.0891e-05 - acc: 1.0000 - binary_crossentropy: 6.0891e-05 - val_loss: 0.7953 - val_acc: 0.8720 - val_binary_crossentropy: 0.7953\n",
      "Epoch 11/20\n",
      " - 11s - loss: 4.7722e-05 - acc: 1.0000 - binary_crossentropy: 4.7722e-05 - val_loss: 0.8107 - val_acc: 0.8716 - val_binary_crossentropy: 0.8107\n",
      "Epoch 12/20\n",
      " - 11s - loss: 3.7963e-05 - acc: 1.0000 - binary_crossentropy: 3.7963e-05 - val_loss: 0.8249 - val_acc: 0.8714 - val_binary_crossentropy: 0.8249\n",
      "Epoch 13/20\n",
      " - 12s - loss: 3.0619e-05 - acc: 1.0000 - binary_crossentropy: 3.0619e-05 - val_loss: 0.8375 - val_acc: 0.8716 - val_binary_crossentropy: 0.8375\n",
      "Epoch 14/20\n",
      " - 11s - loss: 2.5068e-05 - acc: 1.0000 - binary_crossentropy: 2.5068e-05 - val_loss: 0.8501 - val_acc: 0.8716 - val_binary_crossentropy: 0.8501\n",
      "Epoch 15/20\n",
      " - 11s - loss: 2.0688e-05 - acc: 1.0000 - binary_crossentropy: 2.0688e-05 - val_loss: 0.8617 - val_acc: 0.8716 - val_binary_crossentropy: 0.8617\n",
      "Epoch 16/20\n",
      " - 11s - loss: 1.7296e-05 - acc: 1.0000 - binary_crossentropy: 1.7296e-05 - val_loss: 0.8721 - val_acc: 0.8717 - val_binary_crossentropy: 0.8721\n",
      "Epoch 17/20\n",
      " - 11s - loss: 1.4614e-05 - acc: 1.0000 - binary_crossentropy: 1.4614e-05 - val_loss: 0.8833 - val_acc: 0.8715 - val_binary_crossentropy: 0.8833\n",
      "Epoch 18/20\n",
      " - 11s - loss: 1.2470e-05 - acc: 1.0000 - binary_crossentropy: 1.2470e-05 - val_loss: 0.8919 - val_acc: 0.8717 - val_binary_crossentropy: 0.8919\n",
      "Epoch 19/20\n",
      " - 11s - loss: 1.0722e-05 - acc: 1.0000 - binary_crossentropy: 1.0722e-05 - val_loss: 0.9007 - val_acc: 0.8718 - val_binary_crossentropy: 0.9007\n",
      "Epoch 20/20\n",
      " - 11s - loss: 9.3074e-06 - acc: 1.0000 - binary_crossentropy: 9.3074e-06 - val_loss: 0.9094 - val_acc: 0.8717 - val_binary_crossentropy: 0.9094\n"
     ]
    }
   ],
   "source": [
    "bigger_history = bigger_model.fit(train_data, train_labels,\n",
    "                                  epochs=20,\n",
    "                                  batch_size=512,\n",
    "                                  validation_data=(test_data, test_labels),\n",
    "                                  verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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": [
    "def plot_history(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],\n",
    "                   '--', label=name.title()+' Val')\n",
    "    plt.plot(history.epoch, history.history[key], color=val[0].get_color(),\n",
    "             label=name.title()+' Train')\n",
    "\n",
    "  plt.xlabel('Epochs')\n",
    "  plt.ylabel(key.replace('_',' ').title())\n",
    "  plt.legend()\n",
    "\n",
    "  plt.xlim([0,max(history.epoch)])\n",
    "\n",
    "\n",
    "plot_history([('baseline', baseline_history),\n",
    "              ('smaller', smaller_history),\n",
    "              ('bigger', bigger_history)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 25000 samples, validate on 25000 samples\n",
      "Epoch 1/20\n",
      " - 7s - loss: 0.5145 - acc: 0.8151 - binary_crossentropy: 0.4743 - val_loss: 0.3775 - val_acc: 0.8777 - val_binary_crossentropy: 0.3357\n",
      "Epoch 2/20\n",
      " - 6s - loss: 0.3047 - acc: 0.9092 - binary_crossentropy: 0.2592 - val_loss: 0.3340 - val_acc: 0.8880 - val_binary_crossentropy: 0.2858\n",
      "Epoch 3/20\n",
      " - 6s - loss: 0.2563 - acc: 0.9278 - binary_crossentropy: 0.2060 - val_loss: 0.3346 - val_acc: 0.8863 - val_binary_crossentropy: 0.2826\n",
      "Epoch 4/20\n",
      " - 6s - loss: 0.2320 - acc: 0.9388 - binary_crossentropy: 0.1785 - val_loss: 0.3435 - val_acc: 0.8841 - val_binary_crossentropy: 0.2890\n",
      "Epoch 5/20\n",
      " - 6s - loss: 0.2188 - acc: 0.9453 - binary_crossentropy: 0.1628 - val_loss: 0.3563 - val_acc: 0.8802 - val_binary_crossentropy: 0.2994\n",
      "Epoch 6/20\n",
      " - 6s - loss: 0.2063 - acc: 0.9508 - binary_crossentropy: 0.1488 - val_loss: 0.3686 - val_acc: 0.8773 - val_binary_crossentropy: 0.3106\n",
      "Epoch 7/20\n",
      " - 6s - loss: 0.1997 - acc: 0.9526 - binary_crossentropy: 0.1407 - val_loss: 0.3840 - val_acc: 0.8741 - val_binary_crossentropy: 0.3242\n",
      "Epoch 8/20\n",
      " - 6s - loss: 0.1913 - acc: 0.9566 - binary_crossentropy: 0.1309 - val_loss: 0.3905 - val_acc: 0.8734 - val_binary_crossentropy: 0.3298\n",
      "Epoch 9/20\n",
      " - 6s - loss: 0.1862 - acc: 0.9579 - binary_crossentropy: 0.1250 - val_loss: 0.4046 - val_acc: 0.8710 - val_binary_crossentropy: 0.3428\n",
      "Epoch 10/20\n",
      " - 6s - loss: 0.1843 - acc: 0.9589 - binary_crossentropy: 0.1218 - val_loss: 0.4238 - val_acc: 0.8676 - val_binary_crossentropy: 0.3609\n",
      "Epoch 11/20\n",
      " - 6s - loss: 0.1801 - acc: 0.9606 - binary_crossentropy: 0.1166 - val_loss: 0.4279 - val_acc: 0.8676 - val_binary_crossentropy: 0.3638\n",
      "Epoch 12/20\n",
      " - 6s - loss: 0.1767 - acc: 0.9615 - binary_crossentropy: 0.1125 - val_loss: 0.4459 - val_acc: 0.8644 - val_binary_crossentropy: 0.3814\n",
      "Epoch 13/20\n",
      " - 6s - loss: 0.1755 - acc: 0.9627 - binary_crossentropy: 0.1102 - val_loss: 0.4473 - val_acc: 0.8644 - val_binary_crossentropy: 0.3817\n",
      "Epoch 14/20\n",
      " - 6s - loss: 0.1673 - acc: 0.9671 - binary_crossentropy: 0.1017 - val_loss: 0.4566 - val_acc: 0.8620 - val_binary_crossentropy: 0.3913\n",
      "Epoch 15/20\n",
      " - 6s - loss: 0.1628 - acc: 0.9682 - binary_crossentropy: 0.0974 - val_loss: 0.4647 - val_acc: 0.8626 - val_binary_crossentropy: 0.3991\n",
      "Epoch 16/20\n",
      " - 6s - loss: 0.1599 - acc: 0.9694 - binary_crossentropy: 0.0940 - val_loss: 0.4796 - val_acc: 0.8600 - val_binary_crossentropy: 0.4135\n",
      "Epoch 17/20\n",
      " - 6s - loss: 0.1584 - acc: 0.9698 - binary_crossentropy: 0.0923 - val_loss: 0.4817 - val_acc: 0.8600 - val_binary_crossentropy: 0.4151\n",
      "Epoch 18/20\n",
      " - 6s - loss: 0.1558 - acc: 0.9704 - binary_crossentropy: 0.0890 - val_loss: 0.5019 - val_acc: 0.8587 - val_binary_crossentropy: 0.4347\n",
      "Epoch 19/20\n",
      " - 6s - loss: 0.1548 - acc: 0.9706 - binary_crossentropy: 0.0874 - val_loss: 0.5032 - val_acc: 0.8600 - val_binary_crossentropy: 0.4354\n",
      "Epoch 20/20\n",
      " - 6s - loss: 0.1507 - acc: 0.9739 - binary_crossentropy: 0.0826 - val_loss: 0.5163 - val_acc: 0.8574 - val_binary_crossentropy: 0.4480\n"
     ]
    }
   ],
   "source": [
    "l2_model = keras.models.Sequential([\n",
    "    keras.layers.Dense(16, kernel_regularizer=keras.regularizers.l2(0.001),\n",
    "                       activation=tf.nn.relu, input_shape=(NUM_WORDS,)),\n",
    "    keras.layers.Dense(16, kernel_regularizer=keras.regularizers.l2(0.001),\n",
    "                       activation=tf.nn.relu),\n",
    "    keras.layers.Dense(1, activation=tf.nn.sigmoid)\n",
    "])\n",
    "\n",
    "l2_model.compile(optimizer='adam',\n",
    "                 loss='binary_crossentropy',\n",
    "                 metrics=['accuracy', 'binary_crossentropy'])\n",
    "\n",
    "l2_model_history = l2_model.fit(train_data, train_labels,\n",
    "                                epochs=20,\n",
    "                                batch_size=512,\n",
    "                                validation_data=(test_data, test_labels),\n",
    "                                verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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_history([('baseline', baseline_history),\n",
    "              ('l2', l2_model_history)])"
   ]
  },
  {
   "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",
      " - 5s - loss: 0.6173 - acc: 0.6687 - binary_crossentropy: 0.6173 - val_loss: 0.4868 - val_acc: 0.8538 - val_binary_crossentropy: 0.4868\n",
      "Epoch 2/20\n",
      " - 4s - loss: 0.4472 - acc: 0.8210 - binary_crossentropy: 0.4472 - val_loss: 0.3448 - val_acc: 0.8818 - val_binary_crossentropy: 0.3448\n",
      "Epoch 3/20\n",
      " - 4s - loss: 0.3395 - acc: 0.8751 - binary_crossentropy: 0.3395 - val_loss: 0.2872 - val_acc: 0.8888 - val_binary_crossentropy: 0.2872\n",
      "Epoch 4/20\n",
      " - 4s - loss: 0.2748 - acc: 0.9041 - binary_crossentropy: 0.2748 - val_loss: 0.2752 - val_acc: 0.8888 - val_binary_crossentropy: 0.2752\n",
      "Epoch 5/20\n",
      " - 4s - loss: 0.2349 - acc: 0.9216 - binary_crossentropy: 0.2349 - val_loss: 0.2765 - val_acc: 0.8879 - val_binary_crossentropy: 0.2765\n",
      "Epoch 6/20\n",
      " - 4s - loss: 0.2102 - acc: 0.9307 - binary_crossentropy: 0.2102 - val_loss: 0.2860 - val_acc: 0.8852 - val_binary_crossentropy: 0.2860\n",
      "Epoch 7/20\n",
      " - 4s - loss: 0.1753 - acc: 0.9418 - binary_crossentropy: 0.1753 - val_loss: 0.3132 - val_acc: 0.8831 - val_binary_crossentropy: 0.3132\n",
      "Epoch 8/20\n",
      " - 4s - loss: 0.1578 - acc: 0.9480 - binary_crossentropy: 0.1578 - val_loss: 0.3222 - val_acc: 0.8825 - val_binary_crossentropy: 0.3222\n",
      "Epoch 9/20\n",
      " - 4s - loss: 0.1424 - acc: 0.9536 - binary_crossentropy: 0.1424 - val_loss: 0.3465 - val_acc: 0.8809 - val_binary_crossentropy: 0.3465\n",
      "Epoch 10/20\n",
      " - 4s - loss: 0.1279 - acc: 0.9568 - binary_crossentropy: 0.1279 - val_loss: 0.3566 - val_acc: 0.8791 - val_binary_crossentropy: 0.3566\n",
      "Epoch 11/20\n",
      " - 4s - loss: 0.1188 - acc: 0.9611 - binary_crossentropy: 0.1188 - val_loss: 0.3731 - val_acc: 0.8794 - val_binary_crossentropy: 0.3731\n",
      "Epoch 12/20\n",
      " - 4s - loss: 0.1071 - acc: 0.9635 - binary_crossentropy: 0.1071 - val_loss: 0.4026 - val_acc: 0.8786 - val_binary_crossentropy: 0.4026\n",
      "Epoch 13/20\n",
      " - 4s - loss: 0.0966 - acc: 0.9673 - binary_crossentropy: 0.0966 - val_loss: 0.4403 - val_acc: 0.8768 - val_binary_crossentropy: 0.4403\n",
      "Epoch 14/20\n",
      " - 4s - loss: 0.0898 - acc: 0.9677 - binary_crossentropy: 0.0898 - val_loss: 0.4553 - val_acc: 0.8765 - val_binary_crossentropy: 0.4553\n",
      "Epoch 15/20\n",
      " - 4s - loss: 0.0840 - acc: 0.9717 - binary_crossentropy: 0.0840 - val_loss: 0.4696 - val_acc: 0.8762 - val_binary_crossentropy: 0.4696\n",
      "Epoch 16/20\n",
      " - 4s - loss: 0.0765 - acc: 0.9727 - binary_crossentropy: 0.0765 - val_loss: 0.4780 - val_acc: 0.8758 - val_binary_crossentropy: 0.4780\n",
      "Epoch 17/20\n",
      " - 4s - loss: 0.0721 - acc: 0.9738 - binary_crossentropy: 0.0721 - val_loss: 0.5159 - val_acc: 0.8745 - val_binary_crossentropy: 0.5159\n",
      "Epoch 18/20\n",
      " - 5s - loss: 0.0676 - acc: 0.9748 - binary_crossentropy: 0.0676 - val_loss: 0.5141 - val_acc: 0.8747 - val_binary_crossentropy: 0.5141\n",
      "Epoch 19/20\n",
      " - 5s - loss: 0.0667 - acc: 0.9749 - binary_crossentropy: 0.0667 - val_loss: 0.5483 - val_acc: 0.8741 - val_binary_crossentropy: 0.5483\n",
      "Epoch 20/20\n",
      " - 5s - loss: 0.0618 - acc: 0.9749 - binary_crossentropy: 0.0618 - val_loss: 0.5662 - val_acc: 0.8722 - val_binary_crossentropy: 0.5662\n"
     ]
    }
   ],
   "source": [
    "dpt_model = keras.models.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",
    "])\n",
    "\n",
    "dpt_model.compile(optimizer='adam',\n",
    "                  loss='binary_crossentropy',\n",
    "                  metrics=['accuracy','binary_crossentropy'])\n",
    "\n",
    "dpt_model_history = dpt_model.fit(train_data, train_labels,\n",
    "                                  epochs=20,\n",
    "                                  batch_size=512,\n",
    "                                  validation_data=(test_data, test_labels),\n",
    "                                  verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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_history([('baseline', baseline_history),\n",
    "              ('dropout', dpt_model_history)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}