{ "cells": [ { "cell_type": "code", "execution_count": null, "id": "6c180d4b", "metadata": {}, "outputs": [], "source": [ "# Exercise 3\n", "# fashion mnist data\n", "# MLP model with two hidden layers, each with a ReLU activation function.\n", "# Input data is flattened to a 1D array and passed to the model." ] }, { "cell_type": "code", "execution_count": null, "id": "b0e31b9c", "metadata": {}, "outputs": [], "source": [ "import tensorflow as tf\n", "from tensorflow import keras\n", "import matplotlib.pyplot as plt\n", "import numpy as np" ] }, { "cell_type": "code", "execution_count": null, "id": "1ae1412e", "metadata": {}, "outputs": [], "source": [ "# Load the MNIST Fashion dataset\n", "(x_train, y_train), (x_test, y_test) = keras.datasets.fashion_mnist.load_data()" ] }, { "cell_type": "code", "execution_count": null, "id": "f8814914", "metadata": {}, "outputs": [], "source": [ "# Normalize pixel values to between 0 and 1\n", "x_train = x_train.astype(\"float32\") / 255.0\n", "x_test = x_test.astype(\"float32\") / 255.0" ] }, { "cell_type": "code", "execution_count": null, "id": "2810da39", "metadata": {}, "outputs": [], "source": [ "# MNIST dataset images have a shape of (28, 28). The images are flattened\n", "# into a 1D array of length 784 \n", "x_train = x_train.reshape(-1, 784)\n", "x_test = x_test.reshape(-1, 784)" ] }, { "cell_type": "code", "execution_count": null, "id": "96f7ff8a", "metadata": {}, "outputs": [], "source": [ "# The model is defined here with three dense (fully connected) layers\n", "# The first layer is a Dense layer with 128 units and a ReLU activation\n", "# function with an input shape of (784,). This layer serves as the input\n", "# layer of the model.\n", "# The second layer is also a Dense layer with 64 units and a ReLU activation\n", "# function. This layer takes the output of the previous layer as input, and\n", "# applies a non-linear transformation to it to produce a new set of features\n", "# that the next layer can use.\n", "# The third is another Dense layer, one for each class in the output. The\n", "# output is raw scores or logits for each class since there is no activation\n", "# function . This layer is responsible for producing the final output of the\n", "# model, which can then be used to make predictions.\n", "# With Dropout(0.2) 20 % of the input is randomly droped, this should reduce overfitting\n", "model = keras.Sequential([\n", " keras.layers.Dense(128, activation='relu', input_shape=(784,)),\n", " # keras.layers.Dropout(0.2),\n", " keras.layers.Dense(64, activation='relu'),\n", " keras.layers.Dense(10)\n", "])\n" ] }, { "cell_type": "code", "execution_count": null, "id": "a3fe609c", "metadata": {}, "outputs": [], "source": [ "# Compile the model\n", "# adam = specifies the optimizer to use during training\n", "# loss function to use during training, SparseCategoricalCrossentropy loss\n", "# is commonly used for multi-class classification problems.\n", "# from_logits=True indicates that the model's output is a raw score\n", "# for each class and not a probability distribution.\n", "model.compile(optimizer='adam',\n", " loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n", " metrics=['accuracy'])" ] }, { "cell_type": "code", "execution_count": null, "id": "cf6c978d", "metadata": {}, "outputs": [], "source": [ "# Train the model\n", "history = model.fit(x_train, y_train, epochs=10, validation_split=0.2)" ] }, { "cell_type": "code", "execution_count": null, "id": "97fc2313", "metadata": { "scrolled": true }, "outputs": [], "source": [ "# Evaluate the model on the test set\n", "test_loss, test_acc = model.evaluate(x_test, y_test, verbose=2)\n", "print(\"Test accuracy:\", test_acc)" ] }, { "cell_type": "code", "execution_count": null, "id": "ef5f19d0", "metadata": {}, "outputs": [], "source": [ "# Plot the training and validation accuracy and loss over time\n", "plt.figure(figsize=(10, 4))\n", "plt.subplot(1, 2, 1)\n", "plt.plot(history.history[\"accuracy\"])\n", "plt.plot(history.history[\"val_accuracy\"])\n", "plt.title(\"Model accuracy\")\n", "plt.ylabel(\"Accuracy\")\n", "plt.xlabel(\"Epoch\")\n", "plt.legend([\"Train\", \"Validation\"], loc=\"lower right\")\n", "\n", "plt.subplot(1, 2, 2)\n", "plt.plot(history.history[\"loss\"])\n", "plt.plot(history.history[\"val_loss\"])\n", "plt.title(\"Model loss\")\n", "plt.ylabel(\"Loss\")\n", "plt.xlabel(\"Epoch\")\n", "plt.legend([\"Train\", \"Validation\"], loc=\"upper right\")\n", "\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": null, "id": "c0ebddc4", "metadata": {}, "outputs": [], "source": [ "# Plot a confusion matrix of the test set predictions\n", "test_preds = np.argmax(model.predict(x_test), axis=1)\n", "conf_mat = tf.math.confusion_matrix(y_test, test_preds)\n", "plt.imshow(conf_mat, cmap=\"Blues\")\n", "plt.xlabel(\"Predicted labels\")\n", "plt.ylabel(\"True labels\")\n", "plt.xticks(np.arange(10))\n", "plt.yticks(np.arange(10))\n", "plt.colorbar()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": null, "id": "9175d533", "metadata": { "scrolled": true }, "outputs": [], "source": [ "# Make predictions on the test set\n", "y_pred = model.predict(x_test)\n", "y_pred = np.argmax(y_pred, axis=1)\n", "\n", "# Plot some examples from the test set and their predictions\n", "fig, axes = plt.subplots(4, 4, figsize=(18, 18))\n", "for i, ax in enumerate(axes.ravel()):\n", " ax.matshow(x_test[i].reshape(28, 28), cmap='gray')\n", " ax.set_title(\"True: %d\\nPredict: %d\" % (y_test[i], y_pred[i]))\n", " ax.axis(\"off\")\n", "\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": null, "id": "4a6e85be", "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": "3.8.16" } }, "nbformat": 4, "nbformat_minor": 5 }