tf.keras.layers.BatchNormalization

tf.keras.layers.BatchNormalization

1. 批归一化
2. 批标准化

tf.keras.layers.BatchNormalization(
    axis=-1, momentum=0.99, epsilon=0.001, center=True, scale=True,
    beta_initializer='zeros', gamma_initializer='ones',
    moving_mean_initializer='zeros',
    moving_variance_initializer='ones', beta_regularizer=None,
    gamma_regularizer=None, beta_constraint=None, gamma_constraint=None,
    renorm=False, renorm_clipping=None, renorm_momentum=0.99, fused=None,
    trainable=True, virtual_batch_size=None, adjustment=None, name=None, **kwargs
)

keras的这个接口实现的是下面的公式

 

在训练的时候

1. the layer returns (batch - mean(batch)) / (var(batch) + epsilon