Understanding Deep Learning

Excellent book on deep learning. Beautiful graphics. Good mathematics (I wished it was a little more but that’s okay). MIT Press quality.

The book is up-to-date with latest topics in deep learning yet does not require expertise from the reader and gently introduces deep learning to novices. Highly recommended as a textbook or self-study resource. Experienced researchers would still find lots of interesting insights as the book covers a broad range of topics, I certainly learned a few new perspectives!

The AI Bible. No need to say anything else.

This book is a must have for anyone even slightly interested in deep learning.

A great book, but bear in mind the PDF is available free from the website under a Creative Commons licence. So only consider buying the print version (the hard back print version is also excellent). There’s a lot of math in the book, so I expect it to keep me busy for some time!

Nothing further to add to the number of 5 star reviews already available on this page. Up to date and comprehensive, well-written introduction (and beyond) to all aspects of Deep Learning in 2024. If you want a deep understanding of what's going on in the tech behind ChatGPT, facial recognition, automatic translation, etc, you'll do well to find a better overview than this book.

Simon is clearly someone who UNDERSTANDS the material properly, in depth. Look at the explanation for Jensen’s inequality. It’s so well explained — you need at most 5 more minutes reflecting on the explanation after reading it to figure out what the concept is about. So many other good examples of proper scientific writing. It’s better than anything out here, ANYTHING!

I was an A grade A-level maths student 25 years ago, but the calculus I learned then is long forgotten. Given that, this was just too tougher read so didn’t get past the first couple of chapter. Might still come in useful as a reference, but as a professional programmer I do t think I’m going to miss out failing to read this book, not to make use of AI and machine learning tools and libraries out there anyway. If I was going to develop my own tools etc, then I think the first thing to do would be to get back up to speed with A-level or maybe first year undergraduate level pure maths and then think about tackling this book.