Concepts of ANN.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cat detection : Using a single neuron\n",
    "\n",
    "### The  main objective of this project is to explain all the concepts/phases involved in the neural network"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading the packages and dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import h5py\n",
    "import scipy\n",
    "from PIL import Image\n",
    "from scipy import ndimage\n",
    "\n",
    "%matplotlib inline\n",
    "    \n",
    "    \n",
    "def load_dataset():\n",
    "    train_dataset = h5py.File('Cat_dataset/train_catvnoncat.h5', \"r\")\n",
    "    train_set_x_orig = np.array(train_dataset[\"train_set_x\"][:]) # your train set features\n",
    "    train_set_y_orig = np.array(train_dataset[\"train_set_y\"][:]) # your train set labels\n",
    "\n",
    "    test_dataset = h5py.File('Cat_dataset/test_catvnoncat.h5', \"r\")\n",
    "    test_set_x_orig = np.array(test_dataset[\"test_set_x\"][:]) # your test set features\n",
    "    test_set_y_orig = np.array(test_dataset[\"test_set_y\"][:]) # your test set labels\n",
    "\n",
    "    classes = np.array(test_dataset[\"list_classes\"][:]) # the list of classes\n",
    "    \n",
    "    train_set_y_orig = train_set_y_orig.reshape((1, train_set_y_orig.shape[0]))\n",
    "    test_set_y_orig = test_set_y_orig.reshape((1, test_set_y_orig.shape[0]))\n",
    "    \n",
    "    return train_set_x_orig, train_set_y_orig, test_set_x_orig, test_set_y_orig, classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_set_x_orig, train_set_y, test_set_x_orig, test_set_y, classes = load_dataset()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dataset dimensions and its features\n",
    "##### There are 209 images in the train dataset. These are coloured images, hence three dimensional (i.e RGB) and each diemnsion is 64*64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(209, 64, 64, 3)\n",
      "(50, 64, 64, 3)\n",
      "(1, 209)\n",
      "(1, 50)\n"
     ]
    }
   ],
   "source": [
    "print(train_set_x_orig.shape)\n",
    "print(test_set_x_orig.shape)\n",
    "print(train_set_y.shape)\n",
    "print(test_set_y.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting any image from train dataset uisng index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x280efb41f60>"
      ]
     },
     "execution_count": 268,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/A7HnOL/A5Km/kLvJP/HAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x280efae50b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(train_set_x_orig[200])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Flattening the images and then centering them"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_set_x_flattened = train_set_x_orig.reshape(train_set_x_orig.shape[0],-1).T\n",
    "test_set_x_flattened = test_set_x_orig.reshape(test_set_x_orig.shape[0],-1).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(12288, 209)\n",
      "(12288, 50)\n"
     ]
    }
   ],
   "source": [
    "print(train_set_x_flattened.shape)\n",
    "print(test_set_x_flattened.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_X = train_set_x_flattened/255\n",
    "test_X = test_set_x_flattened/255"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setting up the basic functions and parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Defining the activation function\n",
    "def activation_func(z):\n",
    "    sigma = 1/(1+np.exp(-z))\n",
    "    return sigma\n",
    "\n",
    "# Initializing the parameters (weights & bias)\n",
    "def initializing_params(X,bias):\n",
    "    w = np.zeros((X.shape[0],1))\n",
    "    b=bias\n",
    "    return w,b\n",
    "# Forward Propgation through NN to get cost and gradients\n",
    "def forward_prop(w,b,X,Y):\n",
    "    m = X.shape[1]\n",
    "    Z = np.dot(w.transpose(),X) + b       # Calculating w*X +b\n",
    "    A = activation_func(Z)                # passing w*X +b to activation function\n",
    "    cost = -np.sum((Y*np.log(A) + (1-Y)*np.log(1-A)))/m  # Calculating the current cost\n",
    "    dw = np.dot(X,(A-Y).transpose())/m    # Calculating the current gradient of parameter 'w'\n",
    "    db = np.sum((A-Y))/m \n",
    "    \n",
    "    return cost,dw,db\n",
    "\n",
    "# Looping through the NN layer to find optimized parameters (i.e.weights and bias )\n",
    "def optimize_parameters(w, b, X, Y, num_iterations, learning_rate, print_cost):\n",
    "    costs = []\n",
    "    for i in range(num_iterations):\n",
    "        cost,dw,db = forward_prop(w, b, X, Y)\n",
    "        w = w - (learning_rate*dw)\n",
    "        b = b - (learning_rate*db)\n",
    "      \n",
    "        if i % 100 == 0:\n",
    "            costs.append(cost)\n",
    "        \n",
    "        # Print the cost every 100 training iterations\n",
    "        if print_cost and i % 100 == 0:\n",
    "            print (\"Cost after iteration %i: %f\" %(i, cost))\n",
    "    \n",
    "    final_params = {\"w\": w,\n",
    "              \"b\": b}\n",
    "    \n",
    "    Current_grads = {\"dw\": dw,\n",
    "             \"db\": db}\n",
    "    \n",
    "    return final_params, Current_grads, costs\n",
    "\n",
    "# Predict the cases using the final optimized parameters\n",
    "def predict(final_w, final_b, X):\n",
    "    m = X.shape[1]\n",
    "    Y_prediction = np.zeros((1,m))\n",
    "    w = final_w.reshape(X.shape[0], 1)\n",
    "    \n",
    "    # Compute vector \"A\" predicting the probabilities of a cat being present in the picture\n",
    "    A = activation_func(np.dot(final_w.transpose(),X)+final_b)\n",
    "        \n",
    "    for i in range(A.shape[1]):\n",
    "        \n",
    "        # Convert probabilities A[0,i] to actual predictions p[0,i]\n",
    "        if A[0,i]>0.5:\n",
    "            Y_prediction[0,i]=1\n",
    "        else:\n",
    "            Y_prediction[0,i]=0\n",
    "   \n",
    "    return Y_prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### All basic building blocks (functions) are ready. Below chunk will call them in proper order to get the predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cost after iteration 0: 0.693147\n",
      "Cost after iteration 100: 0.584508\n",
      "Cost after iteration 200: 0.466949\n",
      "Cost after iteration 300: 0.376007\n",
      "Cost after iteration 400: 0.331463\n",
      "Cost after iteration 500: 0.303273\n",
      "Cost after iteration 600: 0.279880\n",
      "Cost after iteration 700: 0.260042\n",
      "Cost after iteration 800: 0.242941\n",
      "Cost after iteration 900: 0.228004\n",
      "\n",
      "****  Training set accuracy: 96.65071770334929   ***\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Step 1 : Initializing the weights and bias\n",
    "w,b = initializing_params(train_X,0)\n",
    "\n",
    "# Step 2: Finding the optimized weights and bias after multiple iterations using gradient descent and proper learning rate \n",
    "final_params, Current_grads, costs = optimize_parameters(w, b, train_X, train_set_y, 1000, 0.005, True)\n",
    "\n",
    "# Prediction using the parameters obtained in above step\n",
    "yhat = predict(final_params[\"w\"],final_params[\"b\"],train_X)\n",
    "\n",
    "# Accuracy on train set\n",
    "Accurcay = 100*(1-np.mean(abs(train_set_y-yhat)))\n",
    "print('\\n****  Training set accuracy: ' + str(Accurcay)  + '   ***\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plotting the cost function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x280eb576710>"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x280eb4e0358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(costs)\n",
    "plt.xlabel('Iterations in hundreds')\n",
    "plt.ylabel('Cost')\n",
    "plt.title('Learning rate = 0.005')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Applying the model on test data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "***   Test_accuracy : 72.0  ***\n",
      "\n"
     ]
    }
   ],
   "source": [
    "model_w = final_params['w']\n",
    "model_b= final_params['b']\n",
    "test_yhat = predict(model_w,model_b,test_X)\n",
    "Test_accuracy = 100*(1-np.mean(abs(test_yhat-test_set_y)))\n",
    "print('\\n***   ' + 'Test_accuracy : ' + str(Test_accuracy) + '  ***\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}