Image-Classification-with-CNN-in-TensorFlow.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "91dd4bb2",
   "metadata": {},
   "source": [
    "# Load and Sample the Dataset\n",
    "<p>Your task is to dive deep into the Fashion MNIST dataset and gather insights about the data. Start by loading the dataset. Sample 10% of the training and test sets. After loading, retrieve and display the shape of both the training and test datasets.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d931f809",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import necessary libraries:\n",
    "\n",
    "# Import TensorFlow and NumPy\n",
    "import tensorflow as tf\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "c12d8dce",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
      "29515/29515 [==============================] - 0s 2us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
      "26421880/26421880 [==============================] - 4s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
      "5148/5148 [==============================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
      "4422102/4422102 [==============================] - 1s 0us/step\n"
     ]
    }
   ],
   "source": [
    "#Load the Fashion MNIST dataset:\n",
    "\n",
    "# Load Fashion MNIST dataset\n",
    "fashion_mnist = tf.keras.datasets.fashion_mnist\n",
    "(x_train_total, y_train_total), (x_test_total, y_test_total) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "7aff04f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import the train and test split function from sklearn:\n",
    "\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "10d46fef",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Take only 10% of the data:\n",
    "\n",
    "# Extract a 10% sample from the training dataset\n",
    "x_train, _, y_train, _ = train_test_split(x_train_total, \n",
    "                                          y_train_total, \n",
    "                                          test_size=0.9, \n",
    "                                          random_state=1502)\n",
    " \n",
    "# Extract a 10% sample from the test dataset\n",
    "x_test, _, y_test, _ = train_test_split(x_test_total, \n",
    "                                        y_test_total, \n",
    "                                        test_size=0.9, \n",
    "                                        random_state=1502)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "1d1a6551",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training dataset shape: (6000, 28, 28)\n",
      "Training labels shape: (6000,)\n",
      "Testing dataset shape: (1000, 28, 28)\n",
      "Testing labels shape: (1000,)\n"
     ]
    }
   ],
   "source": [
    "# Print the shapes of the training and testing datasets:\n",
    "\n",
    "# Display the shape of training and testing data\n",
    "print(f\"Training dataset shape: {x_train.shape}\")\n",
    "print(f\"Training labels shape: {y_train.shape}\")\n",
    "print(f\"Testing dataset shape: {x_test.shape}\")\n",
    "print(f\"Testing labels shape: {y_test.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ba1db3ce",
   "metadata": {},
   "source": [
    "# Detailed Visual Analysis of Dataset\n",
    "<div data-purpose=\"safely-set-inner-html:rich-text-viewer:html\" class=\"lab-rich-text-viewer--lab-html-content--22XsM rt-scaffolding\"><p>The Fashion MNIST dataset is a collection of grayscale images of various fashion articles and accessories. The dataset consists of 10 unique categories. Your objective as a data analyst is to provide a detailed visual analysis:</p><p><strong>Random Image Visualization:</strong> For each of the 10 fashion categories (\"T-shirt/top\", \"Trouser\", \"Pullover\", \"Dress\", \"Coat\", \"Sandal\", \"Shirt\", \"Sneaker\", \"Bag\", \"Ankle boot\"), randomly select and display an image. This step is crucial to familiarize yourself with the visual nuances of each category.</p><p><strong>Class Distribution Analysis:</strong> Investigate the distribution of the 10 categories within the dataset using a histogram. This will help determine if the dataset is balanced, meaning it has an equal number of images from each category, or if some categories are over or underrepresented.</p></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "02139744",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the class labels of the Fashion MNIST dataset\n",
    "\n",
    "# Define the class names corresponding to the labels in the dataset\n",
    "class_names = [\"T-shirt/top\", \"Trouser\", \"Pullover\", \"Dress\", \"Coat\",\n",
    "               \"Sandal\", \"Shirt\", \"Sneaker\", \"Bag\", \"Ankle boot\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c3d32456",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Display random images from each of the fashion categories\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    " \n",
    "# Loop through each fashion category\n",
    "for i in range(10):\n",
    "    # Randomly select an image index corresponding to the current fashion category\n",
    "    idx = np.random.choice(np.where(y_train == i)[0])\n",
    "    \n",
    "    # Plot the image\n",
    "    plt.subplot(2, 5, i+1)\n",
    "    plt.imshow(x_train[idx], cmap='gray')\n",
    "    plt.title(class_names[i])\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "94dee762",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1000x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the distribution of fashion categories in the dataset\n",
    "\n",
    "# Create a histogram to display the count of images per fashion category\n",
    "plt.figure(figsize=(10,3))\n",
    "plt.hist(y_train, bins=np.arange(11)-0.5, rwidth=0.8)\n",
    "plt.xticks(np.arange(10), class_names)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f09fab92",
   "metadata": {},
   "source": [
    "# Preprocessing the Dataset for Neural Network Efficiency\n",
    "<p>In this task, the primary objective is to refine and adapt the Fashion MNIST dataset to ensure it's ideally suited for deep learning operations. Begin with a normalization process, transforming the grayscale pixel intensities, which originally range from 0 to 255, to a more neural network-friendly scale of 0-1. Following this, reshape the dataset, adding an essential channel dimension. This step ensures the data aligns perfectly with the structural needs of Convolutional Neural Networks, paving the way for efficient and effective training sessions.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "840929a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalize the training and testing datasets:\n",
    "\n",
    "# Normalize the pixel values to the range [0,1]\n",
    "x_train = x_train / 255.0\n",
    "x_test = x_test / 255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e8789a26",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Adjust the shape of datasets to encompass the channel dimension:\n",
    "\n",
    "# Alter the data shape to be (number of samples, height, width, number of channels)\n",
    "x_train = x_train.reshape((-1, 28, 28, 1))\n",
    "x_test = x_test.reshape((-1, 28, 28, 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d973bb74",
   "metadata": {},
   "source": [
    "# Image Data Augmentation for Model Performance Enhancement\n",
    "<div data-purpose=\"safely-set-inner-html:rich-text-viewer:html\" class=\"lab-rich-text-viewer--lab-html-content--22XsM rt-scaffolding\"><p>To ensure our model can handle various image scenarios and is not restricted to the nuances of our initial dataset, implement a data augmentation process using TensorFlow's <code><strong>ImageDataGenerator</strong></code>. The modifications to be incorporated are:</p><ul><li><p><strong>Rotation:</strong> Up to 10 degrees.</p></li><li><p><strong>Zoom:</strong> A maximum of 10%.</p></li><li><p><strong>Shear:</strong> By 0.1 factor.</p></li><li><p><strong>Width Shift:</strong> Up to 10%.</p></li><li><p><strong>Height Shift:</strong> Up to 10%.</p></li><li><p><strong>Flipping:</strong> Horizontal flipping permitted.</p></li></ul></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d3d88ea0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set up the ImageDataGenerator using the given parameters to alter the image data as specified.\n",
    "\n",
    "# Initializing the ImageDataGenerator with the desired augmentations\n",
    "datagen = tf.keras.preprocessing.image.ImageDataGenerator(\n",
    "   rotation_range=10,      # Rotate images up to 10 degrees\n",
    "   zoom_range=0.1,         # Zoom in on images up to 10%\n",
    "   shear_range=0.1,        # Shear images by a factor of 0.1\n",
    "   width_shift_range=0.1,  # Shift the width of images up to 10%\n",
    "   height_shift_range=0.1, # Shift the height of images up to 10%\n",
    "   horizontal_flip=True    # Allow horizontal flipping of images\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "1d9217d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Implement the defined augmentations to the main training dataset.\n",
    "\n",
    "# Apply the defined data augmentations to the training set\n",
    "datagen.fit(x_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ebb9c7e",
   "metadata": {},
   "source": [
    "# Construct and Train a Convolutional Neural Network (CNN) for Image Classification\n",
    "<div data-purpose=\"safely-set-inner-html:rich-text-viewer:html\" class=\"lab-rich-text-viewer--lab-html-content--22XsM rt-scaffolding\"><p>Develop a Convolutional Neural Network (CNN) for classifying images from the fashion dataset. Your CNN model must be structured as follows:</p><ul><li><p><strong>Initial Convolutional Layer</strong>: 32 filters, a 3x3 kernel, and 'relu' activation.</p></li><li><p><strong>Pooling Layer</strong>: A 2x2 pooling matrix.</p></li><li><p><strong>Dropout Layer</strong>: 20% dropout rate.</p></li><li><p><strong>Second Convolutional Layer</strong>: 64 filters, a 3x3 kernel, and 'relu' activation.</p></li><li><p><strong>Second Pooling Layer</strong>: A 2x2 pooling matrix.</p></li><li><p><strong>Dropout Layer</strong>: 20% dropout rate.</p></li><li><p><strong>Flattening Layer</strong>.</p></li><li><p><strong>Dense Layer</strong>: 256 units with 'relu' activation.</p></li><li><p><strong>Dropout Layer</strong>: 50% dropout rate.</p></li><li><p><strong>Final Dense Layer</strong>: 10 units with 'softmax' activation.</p></li></ul><p>Compile the network using the Adam optimizer, 'sparse_categorical_crossentropy' loss function, and monitor the 'accuracy' metric. Train your model for 5 epochs using a batch size of 32, leveraging the augmented training data and validating against the test dataset.</p><p><br></p></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "a012cc4c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Start by importing the necessary TensorFlow libraries.\n",
    "\n",
    "# Importing required modules from TensorFlow\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense, Dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "ff208754",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Construct the CNN using the specified architecture.\n",
    "\n",
    "# Defining the CNN architecture with the provided specifications\n",
    "model = Sequential([\n",
    "    Conv2D(32, (3,3), activation='relu', input_shape=(28, 28, 1)),  # First convolutional layer\n",
    "    MaxPooling2D(2, 2),                                             # First pooling layer\n",
    "    Dropout(0.2),                                                   # 20% dropout layer\n",
    "    Conv2D(64, (3,3), activation='relu'),                           # Second convolutional layer\n",
    "    MaxPooling2D(2, 2),                                             # Second pooling layer\n",
    "    Dropout(0.2),                                                   # 20% dropout layer\n",
    "    Flatten(),                                                      # Flattening layer\n",
    "    Dense(256, activation='relu'),                                  # Dense layer with 256 units\n",
    "    Dropout(0.5),                                                   # 50% dropout layer\n",
    "    Dense(10, activation='softmax')                                 # Output layer with 10 units\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "52f8b128",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the CNN model with the appropriate parameters.\n",
    "\n",
    "# Compiling the model with the Adam optimizer and suitable loss function\n",
    "model.compile(optimizer='adam',\n",
    "              loss='sparse_categorical_crossentropy',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "7f363802",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-05 00:10:18.053895: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'Placeholder/_0' with dtype int32\n",
      "\t [[{{node Placeholder/_0}}]]\n",
      "2024-02-05 00:10:18.084353: W tensorflow/tsl/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "188/188 [==============================] - 2s 9ms/step - loss: 1.3147 - accuracy: 0.5185 - val_loss: 0.7437 - val_accuracy: 0.7150\n",
      "Epoch 2/5\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.8751 - accuracy: 0.6773 - val_loss: 0.6485 - val_accuracy: 0.7340\n",
      "Epoch 3/5\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.7878 - accuracy: 0.7030 - val_loss: 0.6281 - val_accuracy: 0.7480\n",
      "Epoch 4/5\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.7469 - accuracy: 0.7145 - val_loss: 0.5794 - val_accuracy: 0.7660\n",
      "Epoch 5/5\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.7096 - accuracy: 0.7303 - val_loss: 0.5584 - val_accuracy: 0.7720\n"
     ]
    }
   ],
   "source": [
    "# Proceed to train the model using the augmented dataset.\n",
    "\n",
    "# Training the model with augmented data and validating on the test set\n",
    "history = model.fit(datagen.flow(x_train,\n",
    "                                 y_train,\n",
    "                                 batch_size=32),\n",
    "                    epochs=5,\n",
    "                    validation_data=(x_test,\n",
    "                                     y_test))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72c3a46a",
   "metadata": {},
   "source": [
    "# Implement Advanced Training Techniques for CNN\n",
    "<div data-purpose=\"safely-set-inner-html:rich-text-viewer:html\" class=\"lab-rich-text-viewer--lab-html-content--22XsM rt-scaffolding\"><p>Optimize the training process of your previously constructed Convolutional Neural Network by introducing early stopping and dynamic learning rate adjustments. Implement the following strategies:</p><ul><li><p><strong>Early Stopping</strong>: Monitor the 'val_accuracy' metric and halt training if no improvement is observed after 5 consecutive epochs.</p></li><li><p><strong>Learning Rate Scheduler</strong>: Maintain the current learning rate for the first 10 epochs. For epochs beyond the 10th, reduce the learning rate by a factor of 0.1.</p></li></ul><p>Train your CNN model for a total of 20 epochs, ensuring that you use these callback strategies to halt training early if necessary and dynamically adjust the learning rate during the training process. Note that the number of epochs might change.</p><p><br></p></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "bca1a3ea",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set up the Early Stopping callback.\n",
    "\n",
    "# Implementing early stopping based on validation accuracy\n",
    "early_stop = tf.keras.callbacks.EarlyStopping(monitor='val_accuracy',\n",
    "                                              patience=5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "76bd7a62",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the custom learning rate scheduler function.\n",
    "\n",
    "# Defining the learning rate schedule function\n",
    "def lr_schedule(epoch, lr):\n",
    "    if epoch > 10:\n",
    "        return lr * 0.1  # Reduce learning rate by a factor of 0.1 after 10 epochs\n",
    "    return lr\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "095f2ebf",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initialize the Learning Rate Scheduler callback using the custom function.\n",
    "\n",
    "# Initializing the learning rate scheduler callback\n",
    "scheduler = tf.keras.callbacks.LearningRateScheduler(lr_schedule)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "e5254de5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      " 19/188 [==>...........................] - ETA: 1s - loss: 0.6452 - accuracy: 0.7703"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-02-05 00:11:14.921302: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'Placeholder/_0' with dtype int32\n",
      "\t [[{{node Placeholder/_0}}]]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "188/188 [==============================] - 2s 10ms/step - loss: 0.6735 - accuracy: 0.7402 - val_loss: 0.5393 - val_accuracy: 0.7900 - lr: 0.0010\n",
      "Epoch 2/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.6617 - accuracy: 0.7450 - val_loss: 0.5255 - val_accuracy: 0.7950 - lr: 0.0010\n",
      "Epoch 3/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.6445 - accuracy: 0.7522 - val_loss: 0.5005 - val_accuracy: 0.7980 - lr: 0.0010\n",
      "Epoch 4/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.6138 - accuracy: 0.7585 - val_loss: 0.4828 - val_accuracy: 0.8080 - lr: 0.0010\n",
      "Epoch 5/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.6059 - accuracy: 0.7713 - val_loss: 0.5033 - val_accuracy: 0.7870 - lr: 0.0010\n",
      "Epoch 6/20\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5818 - accuracy: 0.7688 - val_loss: 0.4689 - val_accuracy: 0.8140 - lr: 0.0010\n",
      "Epoch 7/20\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5837 - accuracy: 0.7765 - val_loss: 0.4759 - val_accuracy: 0.8200 - lr: 0.0010\n",
      "Epoch 8/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.5704 - accuracy: 0.7797 - val_loss: 0.4471 - val_accuracy: 0.8260 - lr: 0.0010\n",
      "Epoch 9/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.5626 - accuracy: 0.7858 - val_loss: 0.4496 - val_accuracy: 0.8100 - lr: 0.0010\n",
      "Epoch 10/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.5635 - accuracy: 0.7842 - val_loss: 0.4369 - val_accuracy: 0.8300 - lr: 0.0010\n",
      "Epoch 11/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.5275 - accuracy: 0.7968 - val_loss: 0.4295 - val_accuracy: 0.8360 - lr: 0.0010\n",
      "Epoch 12/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.5310 - accuracy: 0.7993 - val_loss: 0.4056 - val_accuracy: 0.8450 - lr: 1.0000e-04\n",
      "Epoch 13/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.4976 - accuracy: 0.8102 - val_loss: 0.4048 - val_accuracy: 0.8440 - lr: 1.0000e-05\n",
      "Epoch 14/20\n",
      "188/188 [==============================] - 2s 11ms/step - loss: 0.5058 - accuracy: 0.8050 - val_loss: 0.4046 - val_accuracy: 0.8440 - lr: 1.0000e-06\n",
      "Epoch 15/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.4938 - accuracy: 0.8112 - val_loss: 0.4046 - val_accuracy: 0.8440 - lr: 1.0000e-07\n",
      "Epoch 16/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.4894 - accuracy: 0.8115 - val_loss: 0.4046 - val_accuracy: 0.8440 - lr: 1.0000e-08\n",
      "Epoch 17/20\n",
      "188/188 [==============================] - 2s 10ms/step - loss: 0.5063 - accuracy: 0.8055 - val_loss: 0.4046 - val_accuracy: 0.8440 - lr: 1.0000e-09\n"
     ]
    }
   ],
   "source": [
    "# Train the model using both callbacks.\n",
    "\n",
    "# Training the CNN model with augmented data, using the callbacks for early stopping and learning rate scheduling\n",
    "history = model.fit(datagen.flow(x_train,\n",
    "                                 y_train,\n",
    "                                 batch_size=32),\n",
    "                    epochs=20,\n",
    "                    validation_data=(x_test,\n",
    "                                     y_test),\n",
    "                    callbacks=[early_stop, scheduler])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b96fe655",
   "metadata": {},
   "source": [
    "# Evaluate Model Performance and Visualize Misclassifications\n",
    "<div data-purpose=\"safely-set-inner-html:rich-text-viewer:html\" class=\"lab-rich-text-viewer--lab-html-content--22XsM rt-scaffolding\"><p>Evaluate the performance of the trained Convolutional Neural Network on the training and testing datasets. Implement the subsequent evaluation steps:</p><ul><li><p>Plot the accuracy curves for the training and testing datasets.</p></li><li><p>Plot the loss curves for the training and testing datasets.</p></li><li><p>Construct a confusion matrix to analyze the true labels versus the predicted labels for the testing dataset.</p></li><li><p>Visualize the confusion matrix to visually represent misclassifications.</p></li></ul></div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "bcdef7dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Start by importing necessary visualization and matrix evaluation libraries.\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "import seaborn as sns\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "b86239e2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the accuracy curves for both the training and testing data to evaluate the model's performance during training epochs. Use a similar approach to visualize how the model's loss has changed over the training epochs.\n",
    "\n",
    "# Setting up the figure for plotting\n",
    "plt.figure(figsize=(12, 4))\n",
    " \n",
    "# Plotting the accuracy curve for training and testing data\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(history.history['accuracy'], label='Train Accuracy')\n",
    "plt.plot(history.history['val_accuracy'], label='Test Accuracy')\n",
    "plt.legend()\n",
    " \n",
    "# Plotting the loss curve for training and testing data\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(history.history['loss'], label='Train Loss')\n",
    "plt.plot(history.history['val_loss'], label='Test Loss')\n",
    "plt.legend()\n",
    " \n",
    "# Displaying the accuracy plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "0a94d9b0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "32/32 [==============================] - 0s 3ms/step\n"
     ]
    }
   ],
   "source": [
    "# Generate a confusion matrix using the true labels from the test data and the predicted labels from the model.\n",
    "\n",
    "# Predicting labels for the testing dataset\n",
    "y_pred = model.predict(x_test)\n",
    " \n",
    "# Generating the confusion matrix\n",
    "cm = confusion_matrix(y_test, np.argmax(y_pred, axis=1))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "2defc289",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Utilize a heatmap to visually present the confusion matrix, ensuring to include annotations for clarity.\n",
    "\n",
    "# Setting up the figure for the heatmap\n",
    "plt.figure(figsize=(8,8))\n",
    " \n",
    "# Visualizing the confusion matrix using a heatmap\n",
    "sns.heatmap(cm, annot=True, fmt='g', cmap='Blues', xticklabels=class_names, yticklabels=class_names)\n",
    "plt.xlabel('Predicted labels')\n",
    "plt.ylabel('True labels')\n",
    " \n",
    "# Displaying the confusion matrix plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe974307",
   "metadata": {},
   "source": [
    "# Development of a Grad-CAM Heatmap Function for Deep Learning Model Interpretability\n",
    "\n",
    "<p>The goal of this task is to develop and implement a robust function, termed <code><strong>make_gradcam_heatmap</strong></code>, which produces a Grad-CAM heatmap. This heatmap will serve as a visual representation of how much various parts of an input image influenced a deep learning model's predictions. Such visual tools help in understanding and interpreting the decisions made by deep neural networks, particularly in cases where the reasons behind certain predictions need to be clear and justifiable.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "932443df",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Start by naming the function, the inputs and writing the documentation.\n",
    "# Acquire the desired convolutional layer from the model and construct a new model that maps the original input to this layer's activations.\n",
    "# Determine the index or position of our target layer within the model's architecture.\n",
    "# Establish a separate model that processes the activations of our target layer and yields the model's ultimate class predictions.\n",
    "# Utilize TensorFlow's gradient tracking mechanism to monitor the activations and compute the classification scores.\n",
    "# Compute the gradient that represents the importance of each feature in the convolutional layer's activations with respect to the target class score.\n",
    "# Generate the heatmap by weighing the activations by their respective importance, and then normalize it to ensure values lie between 0 and 1.\n",
    "\n",
    "\n",
    "def make_gradcam_heatmap(img_array, model, last_conv_layer_name, pred_index=None):\n",
    "    \"\"\"\n",
    "    Generate a GradCAM heatmap for a given image array.\n",
    "    \n",
    "    Args:\n",
    "    - img_array (numpy.ndarray): The input image array.\n",
    "    - model (tf.keras.Model): The pre-trained model.\n",
    "    - last_conv_layer_name (str): The name of the last convolutional layer in the model.\n",
    "    - pred_index (int, optional): The index of the prediction class. Uses top prediction if None.\n",
    "    \n",
    "    Returns:\n",
    "    - numpy.ndarray: The GradCAM heatmap.\n",
    "    \"\"\"\n",
    "    # Obtain the last convolutional layer from the model\n",
    "    last_conv_layer = model.get_layer(last_conv_layer_name)\n",
    " \n",
    "    # Create a model that maps the model's input to the last_conv_layer's output\n",
    "    last_conv_layer_model = tf.keras.Model(model.inputs, last_conv_layer.output)\n",
    " \n",
    " \n",
    "    # Get the index of the specified convolutional layer in the model\n",
    "    last_conv_layer_idx = [layer.name for layer in model.layers].index(last_conv_layer_name)\n",
    " \n",
    " \n",
    "    # Create an input placeholder with the shape of the output from last_conv_layer\n",
    "    classifier_input = tf.keras.Input(shape=last_conv_layer.output.shape[1:])\n",
    "    x = classifier_input\n",
    " \n",
    "    # Loop through layers after the last_conv_layer and build the classifier model\n",
    "    for layer in model.layers[last_conv_layer_idx+1:]:\n",
    "        x = layer(x)\n",
    "    classifier_model = tf.keras.Model(classifier_input, x)\n",
    " \n",
    "    # Set up the GradientTape for automatic differentiation\n",
    "    with tf.GradientTape() as tape:\n",
    "        # Compute the activations of the last convolutional layer\n",
    "        last_conv_layer_output = last_conv_layer_model(img_array)\n",
    "        # Make sure TensorFlow watches this output for gradient calculations\n",
    "        tape.watch(last_conv_layer_output)\n",
    "        # Predict the class scores using the classifier model\n",
    "        preds = classifier_model(last_conv_layer_output)\n",
    "        # If no specific class index is provided, use the class with the highest score\n",
    "        if pred_index is None:\n",
    "            pred_index = tf.argmax(preds[0])\n",
    "        # Extract the score for the target class\n",
    "        class_channel = preds[:, pred_index]\n",
    " \n",
    "    # Calculate the gradient of the target class score with respect to the activations\n",
    "    grads = tape.gradient(class_channel, last_conv_layer_output)\n",
    " \n",
    "    # Average the gradients spatially (over height and width) to get the importance of each feature map\n",
    "    pooled_grads = tf.reduce_mean(grads, axis=(0, 1, 2))\n",
    " \n",
    " \n",
    "    # Weigh the activations by the importance to get the heatmap\n",
    "    last_conv_layer_output = last_conv_layer_output * pooled_grads[tf.newaxis, tf.newaxis, tf.newaxis, :]\n",
    " \n",
    "    # Average over the feature maps to get a 2D heatmap\n",
    "    heatmap = tf.reduce_mean(last_conv_layer_output, axis=-1)[0]\n",
    "    \n",
    "    # Normalize the heatmap to lie between 0 and 1\n",
    "    heatmap = np.maximum(heatmap, 0) / np.max(heatmap)\n",
    " \n",
    "    return heatmap\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4aaa8fde",
   "metadata": {},
   "source": [
    "# Generate and Display a GradCAM Heatmap Overlay for a Random Test Image\n",
    "<p>For this task, you are to visually represent where the neural network model, specified by the <code><strong>model</strong></code> parameter, is focusing its attention while predicting an image from the test set <code><strong>x_test</strong></code>. The GradCAM technique you'll utilize helps in discerning which regions of an image most influenced the model's prediction.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "8a90a872",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "conv2d\n",
      "max_pooling2d\n",
      "dropout\n",
      "conv2d_1\n",
      "max_pooling2d_1\n",
      "dropout_1\n",
      "flatten\n",
      "dense\n",
      "dropout_2\n",
      "dense_1\n"
     ]
    }
   ],
   "source": [
    "# Enumerate and print the names of all layers in the model.\n",
    "\n",
    "# Iterate over all the layers in the model and print their names\n",
    "for layer in model.layers:\n",
    "    print(layer.name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "5ac7af07",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Randomly pick an image from x_test.\n",
    "\n",
    "# Select a random image from the test dataset\n",
    "img_array = x_test[np.random.choice(x_test.shape[0])]\n",
    " \n",
    "# Reshape the image to add an extra dimension for batch size\n",
    "img_array = img_array[np.newaxis, :]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "3b0f243b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 17ms/step\n"
     ]
    }
   ],
   "source": [
    "# Predict the class of the chosen image.\n",
    "\n",
    "# Get the model's predictions for the selected image\n",
    "preds = model.predict(img_array)\n",
    " \n",
    "# Determine the index of the class with the highest prediction score\n",
    "top_pred_index = np.argmax(preds[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "bb220793",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use the make_gradcam_heatmap function to produce the heatmap.\n",
    "\n",
    "# Generate the GradCAM heatmap for the image based on the specified layer and prediction\n",
    "heatmap = make_gradcam_heatmap(img_array, model, 'conv2d', top_pred_index)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "97889fad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 480x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the heatmap and the original image together.\n",
    "\n",
    "# Display the heatmap\n",
    "plt.matshow(heatmap)\n",
    " \n",
    "# Overlay the original image and heatmap\n",
    "plt.imshow(img_array.reshape(28, 28), cmap='gray', alpha=0.5)\n",
    " \n",
    "# Title the visualization with the model's top prediction\n",
    "plt.title(f\"Predicted class: {class_names[top_pred_index]}\")\n",
    " \n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a05bfb3",
   "metadata": {},
   "source": [
    "# Save, Deploy and Display the Model\n",
    "\n",
    "<p>You are tasked with overseeing the post-training phase of our neural network model. Initiate by ensuring the model's preservation in our designated format, safeguarding both its structure and learned weights. Post storage, retrieve the model into our working environment. Once successfully imported, test its efficacy: select an image at random from our test dataset. Deploy the model for predictions on this image, and subsequently, illustrate the result visually, ensuring it provides clarity on the model's decision-making process. This demonstration will validate our model's robustness and readiness for deployment.</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "b184afa0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Begin by saving the model in the native Keras format.\n",
    "\n",
    "# Save the model\n",
    "model.save(\"cnn_model.keras\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "35de0a6b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the saved model back into the environment.\n",
    "\n",
    "# Load the model\n",
    "loaded_model = tf.keras.models.load_model(\"cnn_model.keras\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "c6fcf27b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Randomly select an image from the test dataset x_test.\n",
    "\n",
    "# Predict on a random test image\n",
    "random_idx = np.random.choice(len(x_test))\n",
    "test_img = x_test[random_idx]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "7350cf48",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 38ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Display the chosen image and overlay with the model's predicted class.\n",
    "\n",
    "# Display the selected image\n",
    "plt.imshow(test_img.reshape(28, 28), cmap='gray')\n",
    " \n",
    "# Predict the class of the image using the loaded model\n",
    "predictions = loaded_model.predict(test_img.reshape(1, 28, 28, 1))\n",
    "predicted_class = np.argmax(predictions)\n",
    " \n",
    "# Add the predicted class as the title\n",
    "plt.title(f\"Predicted: {class_names[predicted_class]}\")\n",
    " \n",
    "# Show the image with the prediction\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "70ce8aba",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}