July 2018
Beginner to intermediate
312 pages
8h 31m
English
Batch normalization rescales the input and all intermediate layer output of the network, making the training smoother and faster. The rescaling is done to achieve a 0 mean and a standard deviation of 1 for all of the input and output. This helps the neural network to train faster and leads to some regularization.
Read now
Unlock full access