Make Your Own Neural Network

This is a great practical introduction to artificial neural networks. Rashid offers a concise explanation of how they are constructed, and summarizes key points at the end of each section. The second half of the book focuses on implementing the techniques described in the first half for building a neural network from scratch in Python, making use of numpy. I read this through twice, and returned to sections on back propagation several more times. For readers who might be turned off by a lot of math, this seems like a good introduction to the topic. For those who need a quick refresher, the author provides an appendix section dedicated to a few basic concepts from Calculus.