Keras: Difference between Kernel and Activity regularizers

Question

I have noticed that weight_regularizer is no more available in Keras and that, in its place, there is activity and kernel regularizer. I would like to know:

What are the main differences between kernel and activity regularizers?

Could I use activity_regularizer in place of weight_regularizer?

Answer 1

We use different kinds of regularizers in deep learning. 

The activity regularizer is used to regulate the output of the neural net. It also helps to regularize hidden layers, so that the output gets optimized. 

The weight regularizer is used to regularize the weights, which means that it decomposes the value of weights so that the neural network can get trained faster.

The new kernel_regularizer replaced the weight_regularizer. It works to regularize the kernel's weight matrix, which affects the overall weights of the neural network.

Hope this answer helps.