Engineering Books

Engineering Books

Listing some technical books that I higly recommend (and actually read).

Deep Learning for Coders with fastai and PyTorch: AI Applications Without a PhD
By Jeremy Howard and Sylvain Gugger

If you've just started your AI journey, this is definitely the first book you should read. This book covers the fundamentals and history of neural network, and continues by creating a basic network using PyTorch and fast ai. It then covers more advanced topics such as different network architectures, how to choose a learning rate and various regularization techniques such as dropout and weight decay.

My favorite episode was the one that showed how to create a neural network from scratch, without using PyTorch internals. I felt that seeing in your own eyes that backpropagation is not some arcane magic makes it all easier.

Programming Massively Parallel Processors: A Hands-on Approach
by Wen-mei W. Hwu, David B. Kirk, and Izzat El Hajj

If you're into AI you have to understand the fundamentals, and by fundamentals I mean - the CUDA platform. This book starts by covering the basics of how the GPU works as opposed to CPU, and will move on to showcase some basic use cases such as matrix multiplication.

Once you've got the fundamental understanding of the CUDA interface and the GPU, it goes further into Neural networks, CNNs and optimization techniques such as tiling and memory coalescing.

Building Recommendation Systems in Python and JAX: Hands-On Production Systems at Scale
by Bryan Bischof Ph.D, Hector Yee

I used to think that recommendation systems are all about collaborative filtering. This was an oversimplification.

This book covers everything you need to know about recommendation systems, including some ways to handle the cold start problem.