if index < max_words:
embedding_vector = embeddings_index.get(word)
if embedding_matrix is not None:
embedding_matrix[index] = embedding_vector
# Build the model
model = models.Sequential()
model.add(layers.Embedding(max_words, embedding_dim, input_length=n_timesteps))
model.add(layers.Flatten())
model.add(layers.Dense(32, activation='relu'))
model.add(layers.Dense(1, activation='sigmoid'))
# Load the pre-trained GloVe embedding in the model
model.layers[0].set_weights([embedding_matrix])
model.layers[0].trainable = False
# Compile
model.compile(optimizer=optimizers.RMSprop(1e-5),
loss='binary_crossentropy',
metrics=['acc'])
history = model.fit(
X_train,
y_train,
epochs=10,
batch_size=32,
validation_data=(X_val, y_val),
)
# Visualization
plot_loss_and_accuracy(history)
#-------------------------------------------------------------------------------
# Modify label dimension for keras
y_2d = to_categorical(y, num_classes=2)
# Build network
model = lab8.keras_nn(n_input=X.shape[1], n_hidden=50, n_output=2)
model.compile(loss='categorical_crossentropy',
optimizer='adam',
metrics=['accuracy'])
# Train network
history = model.fit(X, y_2d, epochs=500, batch_size=10)
# Plot loss and accuracy
utils.plot_loss_and_accuracy(history)
plt.show()
# Plot decision boundary
ax = plt.gca()
utils.plot_decision_boundary(ax,
X,
y,
model,
title='Keras Neural Network Moons',
multiclass=True)
plt.show()
#-------------------------------------------------------------------------------
# Part 7 - Clustering Dataset with Keras
#-------------------------------------------------------------------------------