def create_model(self):
model = Sequential()
model.add(Conv2D(256, (3, 3), input_shape=Resized.shape))
model.add(Activation='relu')
model.Dropout(0.2)
model.add(Conv2D(256. (3, 3)))
model.add(Acitvation='relu')
model.Dropout(0.2)
model.add(Flatten())
model.add(Dense(64))
model.add(Dense(action_choices, activation='linear'))
model.compile(loss='mse',
optimizer=Adam(lr=0.01),
metrics=['accuracy'])
return model
def fit_mlp(X_train, Y_train, X_val, Y_val, neurons, param, verbose=False):
"""
Three hidden layers and one dropout applied to the first hidden layer
"""
mdl = Sequential()
mdl.add(Dense(neurons, input_dim=param.n_lags, activation='tanh'))
mdl.Dropout(param.ratio_dropout)
mdl.add(Dense(neurons, activation='tanh'))
mdl.add(Dense(param.pre_step))
mdl.compile(loss='mean_squared_error', optimizer='adam')
mdl.fit(X_train,
Y_train,
epochs=nb_epochs,
batch_size=nb_batch_size,
verbose=verbose,
shuffle=False)
return mdl
model.add(MaxPool2D(pool_size=(2, 2)))
model.add(Conv2D(32, (2, 2), activation='relu'))
model.add(MaxPool2D(pool_size=(2, 2)))
model.add(Conv2D(64, (2, 2), activation='relu'))
model.add(MaxPool2D(pool_size=(2, 2)))
model.add(Conv2D(32, (2, 2), activation='relu'))
model.add(MaxPool2D(pool_size=(2, 2)))
model.add(Conv2D(64, (2, 2), activation='relu'))
model.add(MaxPool2D(pool_size=(2, 2)))
model.add(Dense(units=1024, activation='relu'))
model.Dropout(0.8)
model.add(Dense(units=2, activation='softmax'))
model.compile(optimizer='adam',
loss='binary_crossentropy',
metrics=['accuracy'])
model.add(Dense(units=1, activation='sigmoid'))
model.compile(optimizer='adam',
loss='binary_crossentropy',
metrics=['accuracy'])
from keras.preprocessing.image import ImageDataGenerator
train_datagen = ImageDataGenerator(rescale=1. / 255,
shear_range=0.2,
zoom_range=0.2,
