Learning Deep Learning: Theory and Practice of Neural Networks, Computer Vision, Natural Language Processing, and Transformers Using TensorFlow

The author did a superb job in explaining the concepts and code using tensorflow api to code a machine learning application using multiple layers (deep learning). Starts from basic concepts like backpropagation, and then in every chapter introduces new deep learning architecture (coded in tensorflow api), together with a working program that applies the architecture to.

This book is so clearly written, it is a joy to read. Whenever I found myself puzzled by something, its explanation would be on the next page. That the book was easy to digest was surprising. What I really didn’t expect was that same “wow” feeling I get from a book by Brian Greene or Daniel Dennett. I do have a EE from a while back so your mileage may vary with the math. Nevertheless, an amazingly lucid work.

This book is so good, I've been feeling guilty for not dropping the author a thank-you note/email. The only book I've read cover-to-cover in two decades. I actually read it twice, and am now making flash cards from the 3rd pass. I recommend this book to all CS (but non-ML) folks and from here you can move on to Chollet's Deep Learning with Python.

I wanted to learn the basic fundamentals of neural networks and am enjoying this book so much I felt like writing a review (rare). I am a generalist and this book is a perfect mixture of dialog, math, and code. Highly recommend this book. Thank you, Magnus Ekman!

I've done this thing the hard way i.e. first learning Stanford CS229, then doing Stanford CS224N and then too, the version of 2019 winter that is available on youtube does not teach you to code Transformers. This book is truly thorough learning guide for Deep Learning, which not only teaches theory but gets your hands dirty on the code. Truly would recommend for any novice new to the field of deep learning.

This book is great for AI practitioners because it presents the recent development of Deep Learning techniques with good amount of codes. It covers wide set of applications from vision to MLP and the depth is just right without going too deep.

If you are a researcher, you should get Goodfellow's Deep Learning book. You will learn all the mathematics and detailed analysis of the algorithms. But if you are an engineer who does not plan to write any paper on this field and just wants to learn it for work, get this book. This books cuts directly to the chase and skips all the mambo-jumbo. It's very fast to read and very entertaining. Which I could not say when I read Goodfellow's book.

Recommended for anyone who wants to enter the world of DL. Basic fundamentals, easy and practiced code and easy to understand.