DeepLearning1.3, Multilayer networks: first steps

views comments

Preparation for the derivation of the Backprop algorithm as gradient descent in layered networks