L-layer application neuron nework week 4.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "5 - L-layer Neural Network\n",
    "Question: Use the helper functions you have implemented previously to build an \n",
    "𝐿\n",
    "L\n",
    "-layer neural network with the following structure: *[LINEAR -> RELU]\n",
    "×\n",
    "×\n",
    "(L-1) -> LINEAR -> SIGMOID*. The functions you may need and their inputs are:\n",
    "def initialize_parameters_deep(layer_dims):\n",
    "    ...\n",
    "    return parameters \n",
    "def L_model_forward(X, parameters):\n",
    "    ...\n",
    "    return AL, caches\n",
    "def compute_cost(AL, Y):\n",
    "    ...\n",
    "    return cost\n",
    "def L_model_backward(AL, Y, caches):\n",
    "    ...\n",
    "    return grads\n",
    "def update_parameters(parameters, grads, learning_rate):\n",
    "    ...\n",
    "    return parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "### CONSTANTS ###\n",
    "layers_dims = [12288, 20, 7, 5, 1] #  5-layer model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def L_layer_model(X, Y, layers_dims, learning_rate=0.0075, num_iterations=3000, print_cost=False): #lr was 0.009\n",
    "    \"\"\"\n",
    "    Implements a L-layer neural network: [LINEAR->RELU]*(L-1)->LINEAR->SIGMOID.\n",
    "    \n",
    "    Arguments:\n",
    "    X -- data, numpy array of shape (number of examples, num_px * num_px * 3)\n",
    "    Y -- true \"label\" vector (containing 0 if cat, 1 if non-cat), of shape (1, number of examples)\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",
    "    \"\"\"\n",
    "\n",
    "    np.random.seed(1)\n",
    "    costs = []                         # keep track of cost\n",
    "    \n",
    "    # Parameters initialization.\n",
    "    ### START CODE HERE ###\n",
    "    parameters = initialize_parameters_deep(layers_dims)\n",
    "    ### END CODE HERE ###\n",
    "    \n",
    "    # Loop (gradient descent)\n",
    "    for i in range(0, num_iterations):\n",
    "\n",
    "        # Forward propagation: [LINEAR -> RELU]*(L-1) -> LINEAR -> SIGMOID.\n",
    "        ### START CODE HERE ### (≈ 1 line of code)\n",
    "        AL, caches = L_model_forward(X, parameters)\n",
    "        ### END CODE HERE ###\n",
    "        \n",
    "        # Compute cost.\n",
    "        ### START CODE HERE ### (≈ 1 line of code)\n",
    "        cost = compute_cost(AL, Y)\n",
    "        ### END CODE HERE ###\n",
    "    \n",
    "        # Backward propagation.\n",
    "        ### START CODE HERE ### (≈ 1 line of code)\n",
    "        grads = L_model_backward(AL, Y, caches)\n",
    "        ### END CODE HERE ###\n",
    " \n",
    "        # Update parameters.\n",
    "        ### START CODE HERE ### (≈ 1 line of code)\n",
    "        parameters = update_parameters(parameters, grads, learning_rate)\n",
    "        ### END CODE HERE ###\n",
    "                \n",
    "        # Print the cost every 100 training example\n",
    "        if print_cost and i % 100 == 0:\n",
    "            print (\"Cost after iteration %i: %f\" % (i, cost))\n",
    "        if print_cost and i % 100 == 0:\n",
    "            costs.append(cost)\n",
    "            \n",
    "    # plot the cost\n",
    "    plt.plot(np.squeeze(costs))\n",
    "    plt.ylabel('cost')\n",
    "    plt.xlabel('iterations (per tens)')\n",
    "    plt.title(\"Learning rate =\" + str(learning_rate))\n",
    "    plt.show()\n",
    "    \n",
    "    return parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "parameters = L_layer_model(train_x, train_y, layers_dims, num_iterations=2500, print_cost=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_train = predict(train_x, train_y, parameters)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_test = predict(test_x, test_y, parameters)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print_mislabeled_images(classes, test_x, test_y, pred_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "## START CODE HERE ##\n",
    "my_image = \"my_image.jpg\" # change this to the name of your image file \n",
    "my_label_y = [1] # the true class of your image (1 -> cat, 0 -> non-cat)\n",
    "## END CODE HERE ##\n",
    "\n",
    "fname = \"images/\" + my_image\n",
    "image = np.array(ndimage.imread(fname, flatten=False))\n",
    "my_image = scipy.misc.imresize(image, size=(num_px,num_px)).reshape((num_px*num_px*3,1))\n",
    "my_predicted_image = predict(my_image, my_label_y, parameters)\n",
    "\n",
    "plt.imshow(image)\n",
    "print (\"y = \" + str(np.squeeze(my_predicted_image)) + \", your L-layer model predicts a \\\"\" + classes[int(np.squeeze(my_predicted_image)),].decode(\"utf-8\") +  \"\\\" picture.\")"
   ]
  }
 ],
 "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}