A must read for anyone looking to get their hands dirty with Tensorflow. This book is practical, succinct and easy to understand. I definitely recommend this to all the M/L enthusiasts who are keen to jump into application oriented learning!
What I liked about this book is it covers the basic techniques such as logistic regression, decision trees, as well as the state-of-the-art like GANs, capsule networks, and deep reinforcement learning. Each chapter is devoted on one concept/technique. Every chapter starts by first explaining intuition and some theory about the concept, and then step-by-step explains how to apply it to solve a practical problem using TensorFlow. Another good thing is that this book covers a wide variety of domains such as "predicting stock prices", "playing pacman", and "sentiment analysis" which gives you a great flavour of how these concepts are used to solve real-world problems.
It's a very lively book filled with charts and figures. It's great for both beginner students as well as practicing data-scientists. Highly recommended!