perceptron.py 文件源码

def __prepare_model(self):
        print('Build model...')
        model = Sequential()
        model.add(TimeDistributedDense(output_dim=self.hidden_cnt,
                                       input_dim=self.input_dim,
                                       input_length=self.input_length,
                                       activation='sigmoid'))
#         model.add(TimeDistributed(Dense(output_dim=self.hidden_cnt,
#                                         input_dim=self.input_dim,
#                                         input_length=self.input_length,
#                                         activation='sigmoid')))
# my modification since import error from keras.layers.core import TimeDistributedMerge
#         model.add(TimeDistributedMerge(mode='ave'))   #comment by me

##################### my ref #########################################################
# # add a layer that returns the concatenation
# # of the positive part of the input and
# # the opposite of the negative part
# 
# def antirectifier(x):
#     x -= K.mean(x, axis=1, keepdims=True)
#     x = K.l2_normalize(x, axis=1)
#     pos = K.relu(x)
#     neg = K.relu(-x)
#     return K.concatenate([pos, neg], axis=1)
# 
# def antirectifier_output_shape(input_shape):
#     shape = list(input_shape)
#     assert len(shape) == 2  # only valid for 2D tensors
#     shape[-1] *= 2
#     return tuple(shape)
# 
# model.add(Lambda(antirectifier, output_shape=antirectifier_output_shape))
#############################################################################

        model.add(Lambda(function=lambda x: K.mean(x, axis=1), 
                   output_shape=lambda shape: (shape[0],) + shape[2:]))
#         model.add(Dropout(0.5))
        model.add(Dropout(0.93755))
        model.add(Dense(self.hidden_cnt, activation='tanh'))
        model.add(Dense(self.output_dim, activation='softmax'))

        # try using different optimizers and different optimizer configs
        print('Compile model...')
#         sgd = SGD(lr=0.1, decay=1e-6, momentum=0.9, nesterov=True)
#         model.compile(loss='categorical_crossentropy', optimizer=sgd)
#         return model
##my add
        adagrad = keras.optimizers.Adagrad(lr=0.01, epsilon=1e-08, decay=0.0)
        model.compile(loss='categorical_crossentropy', optimizer=adagrad)
        return model