The Matrix Calculus You Need For Deep Learning

Terence Parr, Jeremy Howard

This paper is an attempt to explain all the matrix calculus you need in order to understand the training of deep neural networks. We assume no math knowledge beyond what you learne…