autoencoder.py

# Import required libraries

import tensorflow as tf

import numpy as np

import matplotlib.pyplot as plt

# Load MNIST dataset

mnist = tf.keras.datasets.mnist

(x_train, y_train), (x_test, y_test) = mnist.load_data()

# Preprocess data

x_train = x_train.astype('float32') / 255.

x_test = x_test.astype('float32') / 255.

# Define autoencoder model

autoencoder = tf.keras.models.Sequential([

    tf.keras.layers.Flatten(input_shape=(28, 28)),

    tf.keras.layers.Dense(64, activation='relu'),

    tf.keras.layers.Dense(32, activation='relu'),

    tf.keras.layers.Dense(64, activation='relu'),

    tf.keras.layers.Dense(784, activation='sigmoid'),

    tf.keras.layers.Reshape((28, 28))

])

# Compile autoencoder model

autoencoder.compile(optimizer='adam', loss='binary_crossentropy')

# Train autoencoder model

autoencoder.fit(x_train, x_train, epochs=20, batch_size=128, validation_data=(x_test, x_test))

# Evaluate autoencoder model on test set

loss = autoencoder.evaluate(x_test, x_test)

print(f"Autoencoder Loss: {loss}")

# Extract features from encoded layer

encoder = tf.keras.models.Sequential([

    autoencoder.layers[0],

    autoencoder.layers[1],

    autoencoder.layers[2]

])

encoded_train = encoder.predict(x_train)

encoded_test = encoder.predict(x_test)

# Define classifier model

classifier = tf.keras.models.Sequential([

    tf.keras.layers.Dense(256, activation='relu', input_shape=(32,)),

    tf.keras.layers.Dropout(0.5),

    tf.keras.layers.Dense(10, activation='softmax')

])

# Compile classifier model

classifier.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])

# Train classifier model

classifier.fit(encoded_train, y_train, epochs=20, batch_size=128, validation_data=(encoded_test, y_test))

# Evaluate classifier model on test set

loss, accuracy = classifier.evaluate(encoded_test, y_test)

print(f"Classifier Loss: {loss}, Accuracy: {accuracy}")

# Predict labels for test set

predicted_labels = classifier.predict_classes(encoded_test)

# Plot some test images and their predicted labels

fig, axes = plt.subplots(nrows=3, ncols=3)

for i, ax in enumerate(axes.flat):

    ax.imshow(x_test[i], cmap='binary')

    ax.set_title(f"Predicted: {predicted_labels[i]}")

    ax.set_xticks([])

    ax.set_yticks([])

plt.show()