Keras中的连接层
我正试图实现这段代码,作为深度学习的初学者,我无法完全理解他们通过连接层生成“广度和深度神经网络”(WDNN)所做的工作。以下是他们用来生成WDNN的函数:Keras中的连接层,keras,deep-learning,keras-layer,Keras,Deep Learning,Keras Layer,我正试图实现这段代码,作为深度学习的初学者,我无法完全理解他们通过连接层生成“广度和深度神经网络”(WDNN)所做的工作。以下是他们用来生成WDNN的函数: def WDNN(data): input = Input(shape=(data.shape[1],)) x = Dense(256, activation='relu', kernel_regularizer=regularizers.l2(1e-8))(input) x = BatchNormalizatio
def WDNN(data):
input = Input(shape=(data.shape[1],))
x = Dense(256, activation='relu', kernel_regularizer=regularizers.l2(1e-8))(input)
x = BatchNormalization()(x)
x = Dropout(0.5)(x)
x = Dense(256, activation='relu', kernel_regularizer=regularizers.l2(1e-8))(x)
x = BatchNormalization()(x)
x = Dropout(0.5)(x)
x = Dense(256, activation='relu', kernel_regularizer=regularizers.l2(1e-8))(x)
x = BatchNormalization()(x)
x = Dropout(0.5)(x)
wide_deep = concatenate([input, x])
preds = Dense(1, activation='sigmoid', kernel_regularizer=regularizers.l2(1e-8))(wide_deep)
model = Model(input=input, output=preds)
opt = Adam(lr=np.exp(-1.0 * 9))
model.compile(optimizer=opt,
loss='binary_crossentropy',
metrics=['accuracy'])
return model
按照Keras开发者编写的《Keras深度学习》一书中的指导原则,我提出了以下功能。但我无法理解原始函数实际上是如何使用concatenate的,以及如何在自己的代码中实现它以执行相同的操作?如有任何提示,我们将不胜感激
def WDNN(data):
model = models.Sequential()
model.add(layers.Dense(256, activation='relu', kernel_regularizer=regularizers.l2(1e-8), input_shape=(data.shape[1],)))
model.add(layers.BatchNormalization())
model.add(layers.Dropout(0.5))
model.add(layers.Dense(256, activation='relu', kernel_regularizer=regularizers.l2(1e-8)))
model.add(layers.BatchNormalization())
model.add(layers.Dropout(0.5))
model.add(layers.Dense(256, activation='relu', kernel_regularizer=regularizers.l2(1e-8)))
model.add(layers.BatchNormalization())
model.add(layers.Dropout(0.5))
model.add(layers.Dense(1, activation='sigmoid', kernel_regularizer=regularizers.l2(1e-8)))
# Compile model
opt = Adam(lr=np.exp(-1.0 * 9))
model.compile(optimizer=opt,
loss='binary_crossentropy',
metrics=['accuracy'])
return (model)
顺序模型无法实现您的目标。如果要实现
WDNN
,则需要函数模型。请参考。顺序模型无法实现您的目标。如果要实现WDNN
,则需要函数模型。请参阅。