python-kerasHow do I use validation_data when creating a Keras model in Python?

When creating a Keras model in Python, validation_data can be used to evaluate the model's performance on unseen data. Validation_data is a tuple consisting of input data and labels. The input data should be provided in the same format as the training data. The labels should be provided as a one-dimensional array.

Example

# Create a model
model = Sequential()
model.add(Dense(32, input_dim=30))
model.add(Activation('relu'))
model.add(Dense(1))
model.add(Activation('sigmoid'))
model.compile(optimizer='rmsprop',
              loss='binary_crossentropy',
              metrics=['accuracy'])
# Fit the model
model.fit(X_train, y_train, validation_data=(X_val, y_val))

Output example

Train on 8000 samples, validate on 2000 samples
Epoch 1/10
8000/8000 [==============================] - 1s 125us/step - loss: 0.6156 - acc: 0.6778 - val_loss: 0.5307 - val_acc: 0.7495
Epoch 2/10
8000/8000 [==============================] - 0s 57us/step - loss: 0.5138 - acc: 0.7478 - val_loss: 0.4820 - val_acc: 0.7745

Code explanation

• model = Sequential(): This line creates a Sequential model.
• model.add(Dense(32, input_dim=30)): This line adds a Dense layer with 32 nodes and an input dimension of 30.
• model.add(Activation('relu')): This line adds an activation layer with the ReLU activation function.
• model.add(Dense(1)): This line adds a Dense layer with 1 node.
• model.add(Activation('sigmoid')): This line adds an activation layer with the sigmoid activation function.
• model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['accuracy']): This line compiles the model with the RMSprop optimizer, binary crossentropy loss, and accuracy metrics.
• model.fit(X_train, y_train, validation_data=(X_val, y_val)): This line fits the model to the training data with the validation_data tuple.

Helpful links

• Keras Documentation: Model Training
• Keras Documentation: Sequential Model

Edit this code on GitHub