NN_Relu_sigmoid.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.datasets import make_classification\n",
    "from sklearn.model_selection import train_test_split\n",
    "from utils import *\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.colors import ListedColormap\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn import datasets, metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def L_layer_model(X, Y, layers_dims, learning_rate, num_iterations = 3000, print_cost=True):\n",
    "    \"\"\"\n",
    "    Implements a L-layer neural network: [LINEAR->RELU]*(L-1)->LINEAR->SIGMOID.\n",
    "    \n",
    "    Arguments:\n",
    "    X -- data, numpy array of shape (num_px * num_px * 3, number of examples)\n",
    "    Y -- true \"label\" vector \n",
    "    layers_dims -- list containing the input size and each layer size, of length (number of layers + 1).\n",
    "    learning_rate -- learning rate of the gradient descent update rule\n",
    "    num_iterations -- number of iterations of the optimization loop\n",
    "    print_cost -- if True, it prints the cost every 100 steps\n",
    "    \n",
    "    Returns:\n",
    "    parameters -- parameters learnt by the model. They can then be used to predict.\n",
    "    initialize_parameters_deep, compute_cost, update_parameters are loaded from utils\n",
    "    \n",
    "    \"\"\"\n",
    "    \n",
    "    np.random.seed(1)\n",
    "    costs = []                         # keep track of cost\n",
    "    \n",
    "    # Parameters initialization.\n",
    "    \n",
    "    parameters = initialize_parameters_deep(layers_dims)\n",
    "    \n",
    "    # Loop (gradient descent)\n",
    "    for i in range(0, num_iterations):\n",
    "\n",
    "        # Forward propagation: [LINEAR -> RELU]*(L-1) -> LINEAR -> SIGMOID.\n",
    "        \n",
    "        AL, caches = L_model_forward(X, parameters)\n",
    "        \n",
    "        # Compute cost.\n",
    "\n",
    "        cost = compute_cost(AL, Y)\n",
    "        \n",
    "        # Backward propagation.\n",
    "        \n",
    "        grads = L_model_backward(AL, Y, caches)\n",
    "        \n",
    "        # Update parameters.\n",
    "        \n",
    "        parameters = update_parameters(parameters, grads, learning_rate)\n",
    "        \n",
    "        # Print the cost every 100 iterations\n",
    "        if print_cost and i % 100 == 0 or i == num_iterations - 1:\n",
    "            print(\"Cost after iteration {}: {}\".format(i, np.squeeze(cost)))\n",
    "        if i % 100 == 0 or i == num_iterations:\n",
    "            costs.append(cost)\n",
    "    \n",
    "    return parameters, costs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x216 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "digits = datasets.load_digits()\n",
    "\n",
    "_, axes = plt.subplots(nrows=1, ncols=4, figsize=(10, 3))\n",
    "for ax, image, label in zip(axes, digits.images, digits.target):\n",
    "    ax.set_axis_off()\n",
    "    ax.imshow(image, cmap=plt.cm.gray_r, interpolation=\"nearest\")\n",
    "    ax.set_title(\"Training: %i\" % label)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# flatten the images\n",
    "n_samples = len(digits.images)\n",
    "data = digits.images.reshape((n_samples, -1))\n",
    "\n",
    "# Split data into 50% train and 50% test subsets\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    data, digits.target, test_size=0.4, shuffle=False\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10, 1078)\n",
      "(64, 1078)\n"
     ]
    }
   ],
   "source": [
    "\n",
    "m = len(y_train)\n",
    "num_labels = 10\n",
    "\n",
    "y_one_hot = np.zeros((num_labels,m))\n",
    "\n",
    "\n",
    "for i in range(m):\n",
    "    \n",
    "    ind=y_train[i]\n",
    "    y_one_hot[ind,i] = 1\n",
    "\n",
    "Y = y_one_hot\n",
    "X_scaled = X_train/16.\n",
    "X = X_scaled.T\n",
    "\n",
    "print(Y.shape)\n",
    "print(X.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "layers_dims = [64, 25, 10] #  L-layer model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cost after iteration 0: 7.381834214087141\n",
      "Cost after iteration 100: 0.9068161101662283\n",
      "Cost after iteration 200: 0.484468193485973\n",
      "Cost after iteration 300: 0.3464743975204495\n",
      "Cost after iteration 400: 0.2740391147216142\n",
      "Cost after iteration 500: 0.2262265749822302\n",
      "Cost after iteration 600: 0.19183240318645162\n",
      "Cost after iteration 700: 0.1661298576617718\n",
      "Cost after iteration 800: 0.14638475484861838\n",
      "Cost after iteration 900: 0.1307887209591968\n",
      "Cost after iteration 1000: 0.11797009291173949\n",
      "Cost after iteration 1100: 0.10715797985683839\n",
      "Cost after iteration 1200: 0.09784756985232292\n",
      "Cost after iteration 1300: 0.0897181385755805\n",
      "Cost after iteration 1400: 0.08250541185455224\n",
      "Cost after iteration 1500: 0.07603501103664126\n",
      "Cost after iteration 1600: 0.07026639329576387\n",
      "Cost after iteration 1700: 0.06510440660093826\n",
      "Cost after iteration 1800: 0.06047954368907068\n",
      "Cost after iteration 1900: 0.05626436588867452\n",
      "Cost after iteration 1999: 0.05241021770605631\n"
     ]
    }
   ],
   "source": [
    "parameters, costs = L_layer_model(X, Y, layers_dims, learning_rate = 0.2, num_iterations = 2000, print_cost = True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of mislabeled images:  2\n",
      "Accuracy: 99.72\n",
      "List of mislabeled images:  [475, 580]\n",
      "List of mislabeled images:  [475, 580]\n"
     ]
    }
   ],
   "source": [
    "predicted_labels, listofi = predict(X, Y, parameters)\n",
    "print('List of mislabeled images: ', listofi)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "classes = ['Zero', 'One', 'Two', 'Three', 'Four', 'Five', 'Six', 'Seven', 'Eight', 'Nine']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "_, axes = plt.subplots(nrows=1, ncols=2, figsize=(15, 3))\n",
    "for ax, i in zip(axes, listofi):\n",
    "    ax.set_axis_off()\n",
    "    ax.imshow(X[:,i].reshape(8,8), cmap=plt.cm.gray_r, interpolation=\"nearest\")\n",
    "    ax.set_title(\"Prediction: \" + classes[np.argmax(predicted_labels[:,i])] + \" \\n Class: \" + classes[np.argmax(Y[:,i])])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10, 719)\n",
      "(64, 719)\n"
     ]
    }
   ],
   "source": [
    "m = len(y_test)\n",
    "num_labels = 10\n",
    "\n",
    "y_one_hot = np.zeros((num_labels,m))\n",
    "\n",
    "\n",
    "for i in range(m):\n",
    "    \n",
    "    ind=y_test[i]\n",
    "    y_one_hot[ind,i] = 1\n",
    "\n",
    "Y = y_one_hot\n",
    "X_scaled = X_test/16.\n",
    "X = X_scaled.T\n",
    "\n",
    "print(Y.shape)\n",
    "print(X.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of mislabeled images:  2\n",
      "Accuracy: 99.72\n",
      "List of mislabeled images:  [475, 580]\n"
     ]
    }
   ],
   "source": [
    "predicted_labels, listofi = predict(X, Y, parameters)"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}