Keras Deep Learning Cookbook

Adadelta optimizer

It is recommended by the Keras documentation to leave the parameters of this optimizer at their default values.

Let's take a look at the arguments used to initialize this optimizer:

• lr: float >= 0: Learning rate. It's recommended to leave it at the default value, rho: float >= 0.
• epsilon: float >= 0: Fuzz factor. If it is not specified (None), it defaults to K.epsilon().
• decay: float >= 0: Learning rate decay for each update:

model = Sequential()model.add(Dense(512, activation='relu', input_shape=(784,)))model.add(Dropout(0.2))model.add(Dense(512, activation='relu'))model.add(Dropout(0.2))model.add(Dense(num_classes, activation='softmax'))model.summary()ada_delta = keras.optimizers.Adadelta(lr=1.0, rho=0.95, epsilon=None, ...