_notebooks/2020-07-23-02-Improving-Your-Model-Performance.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Improving Your Model Performance\n",
    "> In the previous chapters, you've trained a lot of models! You will now learn how to interpret learning curves to understand your models as they train. You will also visualize the effects of activation functions, batch-sizes, and batch-normalization. Finally, you will learn how to perform automatic hyperparameter optimization to your Keras models using sklearn. This is the Summary of lecture \"Introduction to Deep Learning with Keras\", via datacamp.\n",
    "\n",
    "- toc: true \n",
    "- badges: true\n",
    "- comments: true\n",
    "- author: Chanseok Kang\n",
    "- categories: [Python, Datacamp, Tensorflow-Keras, Deep_Learning]\n",
    "- image: images/val_loss.png"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.rcParams['figure.figsize'] = (8, 8)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Learning curves"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Learning the digits\n",
    "You're going to build a model on the digits dataset, a sample dataset that comes pre-loaded with scikit learn. The digits dataset consist of 8x8 pixel handwritten digits from 0 to 9:\n",
    "![digits](image/digits_dataset_sample.png)\n",
    "You want to distinguish between each of the 10 possible digits given an image, so we are dealing with multi-class classification.\n",
    "The dataset has already been partitioned into `X_train`, `y_train`, `X_test`, and `y_test`, using 30% of the data as testing data. The labels are already one-hot encoded vectors, so you don't need to use Keras `to_categorical()` function.\n",
    "\n",
    "Let's build this new model!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.utils import to_categorical\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X = np.load('./dataset/digits_pixels.npy')\n",
    "y = np.load('./dataset/digits_target.npy')\n",
    "\n",
    "y = to_categorical(y)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[3.2095650e-07 1.8086675e-08 5.8638907e-06 ... 6.3938733e-06\n",
      "  6.9842967e-03 2.7885156e-07]\n",
      " [5.7079029e-07 4.0931591e-05 2.3189692e-02 ... 1.0301257e-03\n",
      "  3.9101571e-01 6.1804475e-04]\n",
      " [2.6313660e-06 2.0355757e-02 1.2905787e-05 ... 7.8606681e-04\n",
      "  2.4483080e-03 5.4859847e-04]\n",
      " ...\n",
      " [2.8147338e-06 4.8216735e-03 6.0069469e-06 ... 6.6422946e-03\n",
      "  1.5737003e-03 1.4798595e-04]\n",
      " [3.2802715e-04 1.2217839e-02 1.7282410e-04 ... 3.1691663e-02\n",
      "  6.9211674e-04 3.9870315e-03]\n",
      " [4.8517606e-05 1.6379059e-05 2.5347804e-03 ... 3.9538622e-02\n",
      "  1.1447263e-01 6.7417235e-03]]\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras import Sequential\n",
    "from tensorflow.keras.layers import Dense\n",
    "\n",
    "# Initialize a Sequential model\n",
    "model = Sequential()\n",
    "\n",
    "# Input and hidden layer with input_shape, 16 neurons, and relu\n",
    "model.add(Dense(16, input_shape=(64, ), activation='relu'))\n",
    "\n",
    "# Output layer with 10 neurons (one per digit) and softmax\n",
    "model.add(Dense(10, activation='softmax'))\n",
    "\n",
    "# Compile your model\n",
    "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
    "\n",
    "# Test if your model is well assembled by predicting before training\n",
    "print(model.predict(X_train))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Predicting on training data inputs before training can help you quickly check that your model works as expected."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Is the model overfitting?\n",
    "Let's train the model you just built and plot its learning curve to check out if it's overfitting! "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_loss(loss,val_loss):\n",
    "    plt.figure();\n",
    "    plt.plot(loss);\n",
    "    plt.plot(val_loss);\n",
    "    plt.title('Model loss');\n",
    "    plt.ylabel('Loss');\n",
    "    plt.xlabel('Epoch');\n",
    "    plt.legend(['Train', 'Test'], loc='upper right');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Train your model for 60 epochs, using X_test and y_test as validation data\n",
    "h_callback = model.fit(X_train, y_train, epochs=60, validation_data=(X_test, y_test), verbose=0)\n",
    "\n",
    "# Extract from the h_callback object loss and val_loss to plot the learning curve\n",
    "plot_loss(h_callback.history['loss'], h_callback.history['val_loss'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This graph doesn't show overfitting but convergence. It looks like your model has learned all it could from the data and it no longer improves. The test loss, although higher than the training loss, is not getting worse, so we aren't overfitting to the training data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Do we need more data?\n",
    "It's time to check whether the digits dataset model you built benefits from more training examples!\n",
    "\n",
    "In order to keep code to a minimum, various things are already initialized and ready to use:\n",
    "\n",
    "- The `model` you just built.\n",
    "- `X_train`,`y_train`,`X_test`, and `y_test`.\n",
    "- The `initial_weights` of your model, saved after using `model.get_weights()`.\n",
    "- A pre-defined list of training sizes: `training_sizes`.\n",
    "- A pre-defined early stopping callback monitoring loss: `early_stop`.\n",
    "- Two empty lists to store the evaluation results: `train_accs` and `test_accs`.\n",
    "\n",
    "Train your model on the different training sizes and evaluate the results on `X_test`. End by plotting the results with plot_results()."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_results(train_accs,test_accs):\n",
    "    plt.plot(training_sizes, train_accs, 'o-', label=\"Training Accuracy\");\n",
    "    plt.plot(training_sizes, test_accs, 'o-', label=\"Test Accuracy\");\n",
    "    plt.title('Accuracy vs Number of training samples');\n",
    "    plt.xlabel('# of training samples');\n",
    "    plt.ylabel('Accuracy');\n",
    "    plt.legend(loc=\"best\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "initial_weights = model.get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 0.0757 - accuracy: 0.9840\n",
      "Epoch 2/50\n",
      "4/4 [==============================] - 0s 982us/step - loss: 0.0693 - accuracy: 0.9840\n",
      "Epoch 3/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0622 - accuracy: 0.9840\n",
      "Epoch 4/50\n",
      "4/4 [==============================] - 0s 966us/step - loss: 0.0574 - accuracy: 0.9840\n",
      "Epoch 5/50\n",
      "4/4 [==============================] - 0s 899us/step - loss: 0.0533 - accuracy: 0.9840\n",
      "Epoch 6/50\n",
      "4/4 [==============================] - 0s 892us/step - loss: 0.0497 - accuracy: 0.9840\n",
      "Epoch 7/50\n",
      "4/4 [==============================] - 0s 872us/step - loss: 0.0458 - accuracy: 0.9920\n",
      "Epoch 8/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0430 - accuracy: 0.9920\n",
      "Epoch 9/50\n",
      "4/4 [==============================] - 0s 902us/step - loss: 0.0395 - accuracy: 1.0000\n",
      "Epoch 10/50\n",
      "4/4 [==============================] - 0s 851us/step - loss: 0.0379 - accuracy: 1.0000\n",
      "Epoch 11/50\n",
      "4/4 [==============================] - 0s 866us/step - loss: 0.0359 - accuracy: 1.0000\n",
      "Epoch 12/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0343 - accuracy: 1.0000\n",
      "Epoch 13/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0330 - accuracy: 1.0000\n",
      "Epoch 14/50\n",
      "4/4 [==============================] - 0s 891us/step - loss: 0.0316 - accuracy: 1.0000\n",
      "Epoch 15/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0301 - accuracy: 1.0000\n",
      "Epoch 16/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0289 - accuracy: 1.0000\n",
      "Epoch 17/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0275 - accuracy: 1.0000\n",
      "Epoch 18/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0268 - accuracy: 1.0000\n",
      "Epoch 19/50\n",
      "4/4 [==============================] - 0s 940us/step - loss: 0.0258 - accuracy: 1.0000\n",
      "Epoch 20/50\n",
      "4/4 [==============================] - 0s 850us/step - loss: 0.0249 - accuracy: 1.0000\n",
      "Epoch 21/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0240 - accuracy: 1.0000\n",
      "Epoch 22/50\n",
      "4/4 [==============================] - 0s 927us/step - loss: 0.0234 - accuracy: 1.0000\n",
      "Epoch 23/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0227 - accuracy: 1.0000\n",
      "Epoch 24/50\n",
      "4/4 [==============================] - 0s 852us/step - loss: 0.0221 - accuracy: 1.0000\n",
      "Epoch 25/50\n",
      "4/4 [==============================] - 0s 791us/step - loss: 0.0216 - accuracy: 1.0000\n",
      "Epoch 26/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0211 - accuracy: 1.0000\n",
      "Epoch 27/50\n",
      "4/4 [==============================] - 0s 859us/step - loss: 0.0205 - accuracy: 1.0000\n",
      "Epoch 28/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0201 - accuracy: 1.0000\n",
      "Epoch 29/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0196 - accuracy: 1.0000\n",
      "Epoch 30/50\n",
      "4/4 [==============================] - 0s 906us/step - loss: 0.0192 - accuracy: 1.0000\n",
      "Epoch 31/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0188 - accuracy: 1.0000\n",
      "Epoch 32/50\n",
      "4/4 [==============================] - 0s 972us/step - loss: 0.0184 - accuracy: 1.0000\n",
      "Epoch 33/50\n",
      "4/4 [==============================] - 0s 935us/step - loss: 0.0180 - accuracy: 1.0000\n",
      "Epoch 34/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0177 - accuracy: 1.0000\n",
      "Epoch 35/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0172 - accuracy: 1.0000\n",
      "Epoch 36/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0170 - accuracy: 1.0000\n",
      "Epoch 37/50\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 0.0167 - accuracy: 1.0000\n",
      "Epoch 38/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0163 - accuracy: 1.0000\n",
      "Epoch 39/50\n",
      "4/4 [==============================] - 0s 918us/step - loss: 0.0161 - accuracy: 1.0000\n",
      "Epoch 40/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0158 - accuracy: 1.0000\n",
      "Epoch 41/50\n",
      "4/4 [==============================] - 0s 859us/step - loss: 0.0156 - accuracy: 1.0000\n",
      "Epoch 42/50\n",
      "4/4 [==============================] - 0s 957us/step - loss: 0.0153 - accuracy: 1.0000\n",
      "Epoch 43/50\n",
      "4/4 [==============================] - 0s 863us/step - loss: 0.0150 - accuracy: 1.0000\n",
      "Epoch 44/50\n",
      "4/4 [==============================] - 0s 894us/step - loss: 0.0148 - accuracy: 1.0000\n",
      "Epoch 45/50\n",
      "4/4 [==============================] - 0s 970us/step - loss: 0.0146 - accuracy: 1.0000\n",
      "Epoch 46/50\n",
      "4/4 [==============================] - 0s 902us/step - loss: 0.0143 - accuracy: 1.0000\n",
      "Epoch 47/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0141 - accuracy: 1.0000\n",
      "Epoch 48/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0139 - accuracy: 1.0000\n",
      "Epoch 49/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0137 - accuracy: 1.0000\n",
      "Epoch 50/50\n",
      "4/4 [==============================] - 0s 1ms/step - loss: 0.0135 - accuracy: 1.0000\n",
      "4/4 [==============================] - 0s 2ms/step - loss: 0.0133 - accuracy: 1.0000\n",
      "17/17 [==============================] - 0s 1ms/step - loss: 0.2311 - accuracy: 0.9370\n",
      "Epoch 1/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0563 - accuracy: 0.9900\n",
      "Epoch 2/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0522 - accuracy: 0.9880\n",
      "Epoch 3/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0487 - accuracy: 0.9900\n",
      "Epoch 4/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0453 - accuracy: 0.9940\n",
      "Epoch 5/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0431 - accuracy: 0.9940\n",
      "Epoch 6/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0411 - accuracy: 0.9940\n",
      "Epoch 7/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0390 - accuracy: 0.9940\n",
      "Epoch 8/50\n",
      "16/16 [==============================] - 0s 999us/step - loss: 0.0379 - accuracy: 0.9940\n",
      "Epoch 9/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0358 - accuracy: 0.9940\n",
      "Epoch 10/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0345 - accuracy: 0.9940\n",
      "Epoch 11/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0334 - accuracy: 0.9960\n",
      "Epoch 12/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0321 - accuracy: 0.9960\n",
      "Epoch 13/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0308 - accuracy: 0.9960\n",
      "Epoch 14/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0303 - accuracy: 0.9960\n",
      "Epoch 15/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0292 - accuracy: 0.9960\n",
      "Epoch 16/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0281 - accuracy: 0.9960\n",
      "Epoch 17/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0270 - accuracy: 0.9960\n",
      "Epoch 18/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0257 - accuracy: 0.9960\n",
      "Epoch 19/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0254 - accuracy: 0.9960\n",
      "Epoch 20/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0244 - accuracy: 0.9980\n",
      "Epoch 21/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0237 - accuracy: 1.0000\n",
      "Epoch 22/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0232 - accuracy: 0.9980\n",
      "Epoch 23/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0223 - accuracy: 1.0000\n",
      "Epoch 24/50\n",
      "16/16 [==============================] - 0s 999us/step - loss: 0.0219 - accuracy: 1.0000\n",
      "Epoch 25/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0211 - accuracy: 1.0000\n",
      "Epoch 26/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0206 - accuracy: 1.0000\n",
      "Epoch 27/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0199 - accuracy: 1.0000\n",
      "Epoch 28/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0194 - accuracy: 1.0000\n",
      "Epoch 29/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0188 - accuracy: 1.0000\n",
      "Epoch 30/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0187 - accuracy: 1.0000\n",
      "Epoch 31/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0181 - accuracy: 1.0000\n",
      "Epoch 32/50\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0175 - accuracy: 1.0000\n",
      "Epoch 33/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0166 - accuracy: 1.0000\n",
      "Epoch 34/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0163 - accuracy: 1.0000\n",
      "Epoch 35/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0157 - accuracy: 1.0000\n",
      "Epoch 36/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0153 - accuracy: 1.0000\n",
      "Epoch 37/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0148 - accuracy: 1.0000\n",
      "Epoch 38/50\n",
      "16/16 [==============================] - 0s 946us/step - loss: 0.0144 - accuracy: 1.0000\n",
      "Epoch 39/50\n",
      "16/16 [==============================] - 0s 971us/step - loss: 0.0139 - accuracy: 1.0000\n",
      "Epoch 40/50\n",
      "16/16 [==============================] - 0s 953us/step - loss: 0.0137 - accuracy: 1.0000\n",
      "Epoch 41/50\n",
      "16/16 [==============================] - 0s 935us/step - loss: 0.0132 - accuracy: 1.0000\n",
      "Epoch 42/50\n",
      "16/16 [==============================] - 0s 958us/step - loss: 0.0131 - accuracy: 1.0000\n",
      "Epoch 43/50\n",
      "16/16 [==============================] - 0s 999us/step - loss: 0.0127 - accuracy: 1.0000\n",
      "Epoch 44/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0124 - accuracy: 1.0000\n",
      "Epoch 45/50\n",
      "16/16 [==============================] - 0s 889us/step - loss: 0.0121 - accuracy: 1.0000\n",
      "Epoch 46/50\n",
      "16/16 [==============================] - 0s 892us/step - loss: 0.0119 - accuracy: 1.0000\n",
      "Epoch 47/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0116 - accuracy: 1.0000\n",
      "Epoch 48/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0114 - accuracy: 1.0000\n",
      "Epoch 49/50\n",
      "16/16 [==============================] - 0s 994us/step - loss: 0.0111 - accuracy: 1.0000\n",
      "Epoch 50/50\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 0.0109 - accuracy: 1.0000\n",
      "16/16 [==============================] - 0s 922us/step - loss: 0.0105 - accuracy: 1.0000\n",
      "17/17 [==============================] - 0s 847us/step - loss: 0.2073 - accuracy: 0.9407\n",
      "Epoch 1/50\n",
      "28/28 [==============================] - 0s 988us/step - loss: 0.0648 - accuracy: 0.9852\n",
      "Epoch 2/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0583 - accuracy: 0.9863\n",
      "Epoch 3/50\n",
      "28/28 [==============================] - 0s 968us/step - loss: 0.0552 - accuracy: 0.9875\n",
      "Epoch 4/50\n",
      "28/28 [==============================] - 0s 948us/step - loss: 0.0520 - accuracy: 0.9863\n",
      "Epoch 5/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0477 - accuracy: 0.9875\n",
      "Epoch 6/50\n",
      "28/28 [==============================] - 0s 969us/step - loss: 0.0454 - accuracy: 0.9909\n",
      "Epoch 7/50\n",
      "28/28 [==============================] - 0s 965us/step - loss: 0.0423 - accuracy: 0.9920\n",
      "Epoch 8/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0418 - accuracy: 0.9909\n",
      "Epoch 9/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0381 - accuracy: 0.9932\n",
      "Epoch 10/50\n",
      "28/28 [==============================] - 0s 974us/step - loss: 0.0365 - accuracy: 0.9898\n",
      "Epoch 11/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0346 - accuracy: 0.9920\n",
      "Epoch 12/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0329 - accuracy: 0.9909\n",
      "Epoch 13/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0308 - accuracy: 0.9943\n",
      "Epoch 14/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0303 - accuracy: 0.9932\n",
      "Epoch 15/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0278 - accuracy: 0.9954\n",
      "Epoch 16/50\n",
      "28/28 [==============================] - 0s 981us/step - loss: 0.0253 - accuracy: 0.9977\n",
      "Epoch 17/50\n",
      "28/28 [==============================] - 0s 939us/step - loss: 0.0248 - accuracy: 0.9977\n",
      "Epoch 18/50\n",
      "28/28 [==============================] - 0s 978us/step - loss: 0.0246 - accuracy: 0.9966\n",
      "Epoch 19/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0231 - accuracy: 0.9977\n",
      "Epoch 20/50\n",
      "28/28 [==============================] - 0s 926us/step - loss: 0.0215 - accuracy: 1.0000\n",
      "Epoch 21/50\n",
      "28/28 [==============================] - 0s 910us/step - loss: 0.0213 - accuracy: 0.9989\n",
      "Epoch 22/50\n",
      "28/28 [==============================] - 0s 919us/step - loss: 0.0198 - accuracy: 1.0000\n",
      "Epoch 23/50\n",
      "28/28 [==============================] - 0s 966us/step - loss: 0.0192 - accuracy: 1.0000\n",
      "Epoch 24/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0182 - accuracy: 1.0000\n",
      "Epoch 25/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0179 - accuracy: 1.0000\n",
      "Epoch 26/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0178 - accuracy: 0.9989\n",
      "Epoch 27/50\n",
      "28/28 [==============================] - 0s 978us/step - loss: 0.0165 - accuracy: 1.0000\n",
      "Epoch 28/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0152 - accuracy: 1.0000\n",
      "Epoch 29/50\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0148 - accuracy: 1.0000\n",
      "Epoch 30/50\n",
      "28/28 [==============================] - 0s 934us/step - loss: 0.0142 - accuracy: 1.0000\n",
      "Epoch 31/50\n",
      "28/28 [==============================] - 0s 1000us/step - loss: 0.0135 - accuracy: 1.0000\n",
      "Epoch 32/50\n",
      "28/28 [==============================] - 0s 963us/step - loss: 0.0142 - accuracy: 1.0000\n",
      "28/28 [==============================] - 0s 1ms/step - loss: 0.0122 - accuracy: 1.0000\n",
      "17/17 [==============================] - 0s 964us/step - loss: 0.1877 - accuracy: 0.9519\n",
      "Epoch 1/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0609 - accuracy: 0.9865\n",
      "Epoch 2/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0608 - accuracy: 0.9849\n",
      "Epoch 3/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0541 - accuracy: 0.9873\n",
      "Epoch 4/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0540 - accuracy: 0.9888\n",
      "Epoch 5/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0475 - accuracy: 0.9920\n",
      "Epoch 6/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0425 - accuracy: 0.9928\n",
      "Epoch 7/50\n",
      "40/40 [==============================] - 0s 977us/step - loss: 0.0420 - accuracy: 0.9920\n",
      "Epoch 8/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0354 - accuracy: 0.9920\n",
      "Epoch 9/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0352 - accuracy: 0.9928\n",
      "Epoch 10/50\n",
      "40/40 [==============================] - 0s 992us/step - loss: 0.0329 - accuracy: 0.9936\n",
      "Epoch 11/50\n",
      "40/40 [==============================] - 0s 953us/step - loss: 0.0310 - accuracy: 0.9952\n",
      "Epoch 12/50\n",
      "40/40 [==============================] - 0s 1ms/step - loss: 0.0288 - accuracy: 0.9952\n",
      "Epoch 13/50\n",
      "40/40 [==============================] - 0s 962us/step - loss: 0.0264 - accuracy: 0.9968\n",
      "Epoch 14/50\n",
      "40/40 [==============================] - 0s 967us/step - loss: 0.0266 - accuracy: 0.9952\n",
      "40/40 [==============================] - 0s 844us/step - loss: 0.0218 - accuracy: 0.9976\n",
      "17/17 [==============================] - 0s 885us/step - loss: 0.1823 - accuracy: 0.9593\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "\n",
    "early_stop = EarlyStopping(monitor='loss', patience=1)\n",
    "train_accs, test_accs = [], []\n",
    "training_sizes = [125, 502, 879, 1255]\n",
    "\n",
    "for size in training_sizes:\n",
    "    # Get a fraction of training data (we only care about the training data)\n",
    "    X_train_frac, y_train_frac = X_train[:size], y_train[:size]\n",
    "    \n",
    "    # Reset the model to the initial weights and train it on the new training data fraction\n",
    "    model.set_weights(initial_weights)\n",
    "    model.fit(X_train_frac, y_train_frac, epochs=50, callbacks=[early_stop])\n",
    "    \n",
    "    # Evaluate and store both: the training data fraction and the complete test set results\n",
    "    train_accs.append(model.evaluate(X_train_frac, y_train_frac)[1])\n",
    "    test_accs.append(model.evaluate(X_test, y_test)[1])\n",
    "    \n",
    "# Plot train vs test accuracies\n",
    "plot_results(train_accs, test_accs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Activation functions\n",
    "![process](image/process.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparing activation functions\n",
    "Comparing activation functions involves a bit of coding, but nothing you can't do!\n",
    "\n",
    "You will try out different activation functions on the multi-label model you built for your farm irrigation machine in chapter 2. The function `get_model('relu')` returns a copy of this model and applies the `'relu'` activation function to its hidden layer.\n",
    "\n",
    "You will loop through several activation functions, generate a new model for each and train it. By storing the history callback in a dictionary you will be able to visualize which activation function performed best in the next exercise!"
   ]
  },
  {
   "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>sensor_0</th>\n",
       "      <th>sensor_1</th>\n",
       "      <th>sensor_2</th>\n",
       "      <th>sensor_3</th>\n",
       "      <th>sensor_4</th>\n",
       "      <th>sensor_5</th>\n",
       "      <th>sensor_6</th>\n",
       "      <th>sensor_7</th>\n",
       "      <th>sensor_8</th>\n",
       "      <th>sensor_9</th>\n",
       "      <th>...</th>\n",
       "      <th>sensor_13</th>\n",
       "      <th>sensor_14</th>\n",
       "      <th>sensor_15</th>\n",
       "      <th>sensor_16</th>\n",
       "      <th>sensor_17</th>\n",
       "      <th>sensor_18</th>\n",
       "      <th>sensor_19</th>\n",
       "      <th>parcel_0</th>\n",
       "      <th>parcel_1</th>\n",
       "      <th>parcel_2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>...</td>\n",
       "      <td>8.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   sensor_0  sensor_1  sensor_2  sensor_3  sensor_4  sensor_5  sensor_6  \\\n",
       "0       1.0       2.0       1.0       7.0       0.0       1.0       1.0   \n",
       "1       5.0       1.0       3.0       5.0       2.0       2.0       1.0   \n",
       "2       3.0       1.0       4.0       3.0       4.0       0.0       1.0   \n",
       "3       2.0       2.0       4.0       3.0       5.0       0.0       3.0   \n",
       "4       4.0       3.0       3.0       2.0       5.0       1.0       3.0   \n",
       "\n",
       "   sensor_7  sensor_8  sensor_9  ...  sensor_13  sensor_14  sensor_15  \\\n",
       "0       4.0       0.0       3.0  ...        8.0        1.0        0.0   \n",
       "1       2.0       3.0       1.0  ...        4.0        5.0        5.0   \n",
       "2       6.0       0.0       2.0  ...        3.0        3.0        1.0   \n",
       "3       2.0       2.0       5.0  ...        4.0        1.0        1.0   \n",
       "4       1.0       1.0       2.0  ...        1.0        3.0        2.0   \n",
       "\n",
       "   sensor_16  sensor_17  sensor_18  sensor_19  parcel_0  parcel_1  parcel_2  \n",
       "0        2.0        1.0        9.0        2.0         0         1         0  \n",
       "1        2.0        2.0        2.0        7.0         0         0         0  \n",
       "2        0.0        3.0        1.0        0.0         1         1         0  \n",
       "3        4.0        1.0        3.0        2.0         0         0         0  \n",
       "4        2.0        1.0        1.0        0.0         1         1         0  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "irrigation = pd.read_csv('./dataset/irrigation_machine.csv', index_col=0)\n",
    "irrigation.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "parcels = irrigation[['parcel_0', 'parcel_1', 'parcel_2']].to_numpy()\n",
    "sensors = irrigation.drop(['parcel_0', 'parcel_1', 'parcel_2'], axis=1).to_numpy()\n",
    "\n",
    "X_train, X_test, y_train, y_test = \\\n",
    "    train_test_split(sensors, parcels, test_size=0.3, stratify=parcels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(1)\n",
    "\n",
    "# Return a new model with the given activation\n",
    "def get_model(act_function):\n",
    "    model = Sequential()\n",
    "    if act_function == 'leaky_relu':\n",
    "        model.add(Dense(64, input_shape=(20, ), activation=tf.nn.leaky_relu))\n",
    "    else:\n",
    "        model.add(Dense(64, input_shape=(20, ), activation=act_function))\n",
    "\n",
    "    # Add an output layer of 3 neurons with sigmoid activation\n",
    "    model.add(Dense(3, activation='sigmoid'))\n",
    "    # Compile your model with binary crossentropy loss\n",
    "    model.compile(optimizer='adam',\n",
    "              loss='binary_crossentropy',\n",
    "              metrics=['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finishing with relu...\n",
      "Finishing with leaky_relu...\n",
      "Finishing with sigmoid...\n",
      "Finishing with tanh...\n"
     ]
    }
   ],
   "source": [
    "# Activation functions to try\n",
    "activations = ['relu', 'leaky_relu', 'sigmoid', 'tanh']\n",
    "\n",
    "# Loop over the activation functions\n",
    "activation_results = {}\n",
    "\n",
    "for act in activations:\n",
    "    # Get a new model with the current activation\n",
    "    model = get_model(act)\n",
    "    \n",
    "    # Fit the model and store the history results\n",
    "    h_callback = model.fit(X_train, y_train, epochs=100, validation_data=(X_test, y_test), verbose=0)\n",
    "    activation_results[act] = h_callback\n",
    "    print('Finishing with {}...'.format(act))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparing activation functions II\n",
    "For every `h_callback` of each activation function in activation_results:\n",
    "\n",
    "- `The h_callback.history['val_loss']` has been extracted.\n",
    "- `The h_callback.history['val_acc']` has been extracted."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "val_loss_per_function = {}\n",
    "val_acc_per_function = {}\n",
    "\n",
    "for k, v in activation_results.items():\n",
    "    val_loss_per_function[k] = v.history['val_loss']\n",
    "    val_acc_per_function[k] = v.history['val_accuracy']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a dataframe from val_loss_per_function\n",
    "val_loss = pd.DataFrame(val_loss_per_function)\n",
    "\n",
    "# Call plot on the dataframe\n",
    "val_loss.plot(title='validation Loss');\n",
    "\n",
    "# Create a dataframe from val_acc_per_function\n",
    "val_acc = pd.DataFrame(val_acc_per_function)\n",
    "\n",
    "# Call plot on the dataframe\n",
    "val_acc.plot(title='validation Accuracy');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You've plotted both: loss and accuracy curves. It looks like sigmoid activation worked best for this particular model as the hidden layer's activation function. It led to a model with lower validation loss and higher accuracy after 100 epochs."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Batch size and batch normalization\n",
    "- Mini-batches\n",
    "    - Advantages\n",
    "        - Networks train faster (more weight updates in same amount of time)\n",
    "        - Less RAM memory required, can train on huge datasets\n",
    "        - Noise can help networks reach a lower error, escaping local minima\n",
    "    - Disadvantage\n",
    "        - More iterations need to be run\n",
    "        - Need to be adjusted, we need to find a good batch size\n",
    "- Batch Normalization\n",
    "    - Advantages\n",
    "        - Improves gradient flow\n",
    "        - Allows higher learning rates\n",
    "        - Reduces dependence on weight initializations\n",
    "        - Acts as an unintended form of regularization\n",
    "        - Limits inter covariate shift"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Changing batch sizes\n",
    "You've seen models are usually trained in batches of a fixed size. The smaller a batch size, the more weight updates per epoch, but at a cost of a more unstable gradient descent. Specially if the batch size is too small and it's not representative of the entire training set.\n",
    "\n",
    "Let's see how different batch sizes affect the accuracy of a simple binary classification model that separates red from blue dots.\n",
    "\n",
    "You'll use a batch size of one, updating the weights once per sample in your training set for each epoch. Then you will use the entire dataset, updating the weights only once per epoch."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.242655</td>\n",
       "      <td>0.038320</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.044330</td>\n",
       "      <td>-0.056673</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.786777</td>\n",
       "      <td>-0.757186</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.004067</td>\n",
       "      <td>0.131172</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.164107</td>\n",
       "      <td>0.150650</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1  2\n",
       "0  0.242655  0.038320  1\n",
       "1  0.044330 -0.056673  1\n",
       "2 -0.786777 -0.757186  0\n",
       "3  0.004067  0.131172  1\n",
       "4 -0.164107  0.150650  1"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dots = pd.read_csv('./dataset/dots.csv')\n",
    "dots.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = dots.iloc[:, :-1]\n",
    "y = dots.iloc[:, -1]\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_model():\n",
    "    model = Sequential()\n",
    "    model.add(Dense(4,input_shape=(2,),activation='relu'))\n",
    "    model.add(Dense(1,activation=\"sigmoid\"))\n",
    "    model.compile('sgd', 'binary_crossentropy', metrics=['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "700/700 [==============================] - 1s 908us/step - loss: 0.6371 - accuracy: 0.5986\n",
      "Epoch 2/5\n",
      "700/700 [==============================] - 1s 897us/step - loss: 0.5491 - accuracy: 0.8343\n",
      "Epoch 3/5\n",
      "700/700 [==============================] - 1s 859us/step - loss: 0.4739 - accuracy: 0.8786\n",
      "Epoch 4/5\n",
      "700/700 [==============================] - 1s 854us/step - loss: 0.4095 - accuracy: 0.9014\n",
      "Epoch 5/5\n",
      "700/700 [==============================] - 1s 878us/step - loss: 0.3501 - accuracy: 0.9157\n",
      "10/10 [==============================] - 0s 803us/step - loss: 0.3399 - accuracy: 0.9200\n",
      "\n",
      " The accuracy when using a batch of size 1 is:  0.9200000166893005\n"
     ]
    }
   ],
   "source": [
    "# Get a fresh new model with get_model\n",
    "model = get_model()\n",
    "\n",
    "# Train your model for 5 epochs with a batch size of 1\n",
    "model.fit(X_train, y_train, epochs=5, batch_size=1)\n",
    "print(\"\\n The accuracy when using a batch of size 1 is: \", model.evaluate(X_test, y_test)[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "1/1 [==============================] - 0s 751us/step - loss: 0.6923 - accuracy: 0.5357\n",
      "Epoch 2/5\n",
      "1/1 [==============================] - 0s 622us/step - loss: 0.6921 - accuracy: 0.5357\n",
      "Epoch 3/5\n",
      "1/1 [==============================] - 0s 519us/step - loss: 0.6919 - accuracy: 0.5357\n",
      "Epoch 4/5\n",
      "1/1 [==============================] - 0s 515us/step - loss: 0.6917 - accuracy: 0.5314\n",
      "Epoch 5/5\n",
      "1/1 [==============================] - 0s 475us/step - loss: 0.6915 - accuracy: 0.5286\n",
      "10/10 [==============================] - 0s 828us/step - loss: 0.6952 - accuracy: 0.4367\n",
      "\n",
      " The accuracy when using the whole training set as batch-size was:  0.43666666746139526\n"
     ]
    }
   ],
   "source": [
    "model = get_model()\n",
    "\n",
    "# Fit your model for 5 epochs with a batch of size the training set\n",
    "model.fit(X_train, y_train, epochs=5, batch_size=X_train.shape[0])\n",
    "print(\"\\n The accuracy when using the whole training set as batch-size was: \",\n",
    "      model.evaluate(X_test, y_test)[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can see that accuracy is lower when using a batch size equal to the training set size. This is not because the network had more trouble learning the optimization function: Even though the same number of epochs were used for both batch sizes the number of resulting weight updates was very different!. With a batch of size the training set and 5 epochs we only get 5 updates total, each update computes and averaged gradient descent with all the training set observations. To obtain similar results with this batch size we should increase the number of epochs so that more weight updates take place."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Batch normalizing a familiar model\n",
    "Remember the digits dataset you trained in the first exercise of this chapter?\n",
    "\n",
    "![digits](image/digits_dataset_sample.png)\n",
    "\n",
    "A multi-class classification problem that you solved using `softmax` and 10 neurons in your output layer.\n",
    "You will now build a new deeper model consisting of 3 hidden layers of 50 neurons each, using batch normalization in between layers. The `kernel_initializer` parameter is used to initialize weights in a similar way."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.layers import BatchNormalization\n",
    "\n",
    "# Build your deep network\n",
    "batchnorm_model = Sequential()\n",
    "batchnorm_model.add(Dense(50, input_shape=(64, ), activation='relu', kernel_initializer='normal'))\n",
    "batchnorm_model.add(BatchNormalization())\n",
    "batchnorm_model.add(Dense(50, activation='relu', kernel_initializer='normal'))\n",
    "batchnorm_model.add(BatchNormalization())\n",
    "batchnorm_model.add(Dense(50, activation='relu', kernel_initializer='normal'))\n",
    "batchnorm_model.add(BatchNormalization())\n",
    "batchnorm_model.add(Dense(10, activation='softmax', kernel_initializer='normal'))\n",
    "\n",
    "# Compile your model with sgd\n",
    "batchnorm_model.compile(optimizer='sgd', loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Batch normalization effects\n",
    "Batch normalization tends to increase the learning speed of our models and make their learning curves more stable. Let's see how two identical models with and without batch normalization compare.\n",
    "\n",
    "The model you just built `batchnorm_model` is loaded for you to use. An exact copy of it without batch normalization: `standard_model`, is available as well.\n",
    "\n",
    "You will compare the accuracy learning curves for both models plotting them with compare_histories_acc()."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compare_histories_acc(h1,h2):\n",
    "    plt.plot(h1.history['accuracy']);\n",
    "    plt.plot(h1.history['val_accuracy']);\n",
    "    plt.plot(h2.history['accuracy']);\n",
    "    plt.plot(h2.history['val_accuracy']);\n",
    "    plt.title(\"Batch Normalization Effects\");\n",
    "    plt.xlabel('Epoch');\n",
    "    plt.ylabel('Accuracy');\n",
    "    plt.legend(['Train', 'Test', 'Train with Batch Normalization', 'Test with Batch Normalization'], loc='best');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = np.load('./dataset/digits_pixels.npy')\n",
    "y = np.load('./dataset/digits_target.npy')\n",
    "\n",
    "y = to_categorical(y)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build your deep network\n",
    "standard_model = Sequential()\n",
    "standard_model.add(Dense(50, input_shape=(64, ), activation='relu', kernel_initializer='normal'))\n",
    "standard_model.add(Dense(50, activation='relu', kernel_initializer='normal'))\n",
    "standard_model.add(Dense(50, activation='relu', kernel_initializer='normal'))\n",
    "standard_model.add(Dense(10, activation='softmax', kernel_initializer='normal'))\n",
    "\n",
    "# Compile your model with sgd\n",
    "standard_model.compile(optimizer='sgd', loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Train your standard model, storing its history callback\n",
    "h1_callback = standard_model.fit(X_train, y_train, validation_data=(X_test, y_test), epochs=10, verbose=0)\n",
    "\n",
    "# Train the batch normalized model you recently built, store its history callback\n",
    "h2_callback = batchnorm_model.fit(X_train, y_train, validation_data=(X_test, y_test), epochs=10, verbose=0)\n",
    "\n",
    "# Call compare_histories_acc passing in both model histories\n",
    "compare_histories_acc(h1_callback, h2_callback)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can see that for this deep model batch normalization proved to be useful, helping the model obtain high accuracy values just over the first 10 training epochs."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hyperparameter tuning\n",
    "- Neural network hyperparameters\n",
    "    - Number of layers\n",
    "    - Number of neurons per layer\n",
    "    - Layer order\n",
    "    - Layer activations\n",
    "    - Batch sizes\n",
    "    - Learning rates\n",
    "    - Optimizers\n",
    "    - ...\n",
    "- Tips for neural networks hyperparameter tuning\n",
    "    - Random search is preferred over grid search\n",
    "    - Don't use many epochs\n",
    "    - Use a smaller sample of your dataset\n",
    "    - Play with batch sizes, activations, optimizers and learning rates"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preparing a model for tuning\n",
    "Let's tune the hyperparameters of a binary classification model that does well classifying the breast cancer dataset.\n",
    "\n",
    "You've seen that the first step to turn a model into a sklearn estimator is to build a function that creates it. The definition of this function is important since hyperparameter tuning is carried out by varying the arguments your function receives.\n",
    "\n",
    "Build a simple `create_model()` function that receives both a learning rate and an activation function as arguments.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creates a model given an activation and learning rate\n",
    "def create_model(learning_rate=0.01, activation='relu'):\n",
    "    # Create an Adam optimizer with the given learning rate\n",
    "    opt = tf.keras.optimizers.Adam(lr=learning_rate)\n",
    "    \n",
    "    # Create your binary classification model\n",
    "    model = Sequential()\n",
    "    model.add(Dense(128, input_shape=(30, ), activation=activation))\n",
    "    model.add(Dense(256, activation=activation))\n",
    "    model.add(Dense(1, activation='sigmoid'))\n",
    "    \n",
    "    # Compile your model with your optimizer, loss and metrics\n",
    "    model.compile(optimizer=opt, loss='binary_crossentropy', metrics=['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Tuning the model parameters\n",
    "It's time to try out different parameters on your model and see how well it performs!\n",
    "\n",
    "The `create_model()` function you built in the previous exercise is ready for you to use.\n",
    "\n",
    "Since fitting the `RandomizedSearchCV` object would take too long, the results you'd get are printed in the `show_results()` function. You could try `random_search.fit(X,y)` in the console yourself to check it does work after you have built everything else, but you will probably timeout the exercise (so copy your code first if you try this or you can lose your progress!).\n",
    "\n",
    "You don't need to use the optional `epochs` and `batch_size` parameters when building your `KerasClassifier` object since you are passing them as params to the random search and this works already."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_breast_cancer\n",
    "\n",
    "X, y = load_breast_cancer(return_X_y=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6/6 [==============================] - 0s 836us/step - loss: 0.2679 - accuracy: 0.8579\n",
      "6/6 [==============================] - 0s 978us/step - loss: 0.2163 - accuracy: 0.9000\n",
      "6/6 [==============================] - 0s 876us/step - loss: 0.4303 - accuracy: 0.8042\n",
      "2/2 [==============================] - 0s 893us/step - loss: 0.8017 - accuracy: 0.4895\n",
      "2/2 [==============================] - 0s 920us/step - loss: 0.8923 - accuracy: 0.6211\n",
      "2/2 [==============================] - 0s 840us/step - loss: 0.6281 - accuracy: 0.7725\n",
      "6/6 [==============================] - 0s 972us/step - loss: 0.7374 - accuracy: 0.5105\n",
      "6/6 [==============================] - 0s 988us/step - loss: 0.7063 - accuracy: 0.3789\n",
      "6/6 [==============================] - 0s 1ms/step - loss: 0.5781 - accuracy: 0.7725\n",
      "1/1 [==============================] - 0s 643us/step - loss: 0.6987 - accuracy: 0.5105\n",
      "1/1 [==============================] - 0s 604us/step - loss: 0.9258 - accuracy: 0.3789\n",
      "1/1 [==============================] - 0s 756us/step - loss: 0.6346 - accuracy: 0.7725\n",
      "6/6 [==============================] - 0s 863us/step - loss: 0.6929 - accuracy: 0.5105\n",
      "6/6 [==============================] - 0s 992us/step - loss: 1.3970 - accuracy: 0.3789\n",
      "6/6 [==============================] - 0s 822us/step - loss: 0.7859 - accuracy: 0.2275\n",
      "2/2 [==============================] - 0s 1ms/step - loss: 0.3839 - accuracy: 0.8842\n",
      "2/2 [==============================] - 0s 886us/step - loss: 0.2792 - accuracy: 0.9053\n",
      "2/2 [==============================] - 0s 845us/step - loss: 0.3960 - accuracy: 0.8519\n",
      "1/1 [==============================] - 0s 628us/step - loss: 0.2547 - accuracy: 0.8947\n",
      "1/1 [==============================] - 0s 752us/step - loss: 0.2135 - accuracy: 0.9105\n",
      "1/1 [==============================] - 0s 623us/step - loss: 0.2748 - accuracy: 0.9101\n",
      "6/6 [==============================] - 0s 871us/step - loss: 0.8303 - accuracy: 0.4895\n",
      "6/6 [==============================] - 0s 852us/step - loss: 0.3743 - accuracy: 0.8947\n",
      "6/6 [==============================] - 0s 872us/step - loss: 0.5416 - accuracy: 0.7725\n",
      "2/2 [==============================] - 0s 891us/step - loss: 0.7883 - accuracy: 0.4895\n",
      "2/2 [==============================] - 0s 866us/step - loss: 0.6587 - accuracy: 0.6263\n",
      "2/2 [==============================] - 0s 964us/step - loss: 0.6417 - accuracy: 0.7725\n",
      "2/2 [==============================] - 0s 805us/step - loss: 0.3068 - accuracy: 0.8895\n",
      "2/2 [==============================] - 0s 942us/step - loss: 0.6565 - accuracy: 0.3789\n",
      "2/2 [==============================] - 0s 854us/step - loss: 0.2880 - accuracy: 0.9206\n",
      "Best: 0.9051053524017334 using {'learning_rate': 0.01, 'epochs': 50, 'batch_size': 256, 'activation': 'relu'}\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.wrappers.scikit_learn import KerasClassifier\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "# Create a KerasClassifier\n",
    "model = KerasClassifier(build_fn=create_model)\n",
    "\n",
    "# Define the parameters to try out\n",
    "params = {\n",
    "    'activation': ['relu', 'tanh'],\n",
    "    'batch_size': [32, 128, 256],\n",
    "    'epochs': [50, 100, 200],\n",
    "    'learning_rate': [0.1, 0.01, 0.001]\n",
    "}\n",
    "\n",
    "# Create a randomize search cv object passing in the parameters to try out\n",
    "random_search = RandomizedSearchCV(model, param_distributions=params, cv=KFold(3))\n",
    "random_search_results = random_search.fit(X, y, verbose=0)\n",
    "\n",
    "# Print results\n",
    "print(\"Best: {} using {}\".format(random_search_results.best_score_,\n",
    "                                 random_search_results.best_params_))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training with cross-validation\n",
    "Time to train your model with the best parameters found: 0.01 for the learning rate, 100 epochs, a 128 batch_size and `relu` activations.\n",
    "\n",
    "The `create_model()` function from the previous exercise is ready for you to use. `X` and `y` are loaded as features and labels.\n",
    "\n",
    "Use the best values found for your model when creating your `KerasClassifier` object so that they are used when performing cross_validation.\n",
    "\n",
    "End this chapter by training an awesome tuned model on the breast cancer dataset!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The mean accuracy was:  0.9261858264605204\n",
      "With a standard deviation of:  0.00430127209155373\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "# Create a KerasClassifier\n",
    "model = KerasClassifier(build_fn=create_model, \n",
    "                        epochs=100, batch_size=128, verbose=0)\n",
    "\n",
    "# Calculate the accuracy score for each fold\n",
    "kfolds = cross_val_score(model, X, y, cv=3)\n",
    "\n",
    "# Print the mean accuracy\n",
    "print('The mean accuracy was: ', kfolds.mean())\n",
    "\n",
    "# Print the accuracy standard deviation\n",
    "print('With a standard deviation of: ', kfolds.std())"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}