Reinforcement learning (RL) will deliver one of the biggest breakthroughs in AI over the next decade, enabling algorithms to learn from their environment to achieve arbitrary goals. This exciting development avoids constraints found in traditional machine learning (ML) algorithms. This practical book shows data science and AI professionals how to learn by reinforcement and enable a machine to learn by itself.
Author Phil Winder of Winder Research covers everything from basic building blocks to state-of-the-art practices. You'll explore the current state of RL, focus on industrial applications, learn numerous algorithms, and benefit from dedicated chapters on deploying RL solutions to production. This is no cookbook; doesn't shy away from math and expects familiarity with ML.
Learn what RL is and how the algorithms help solve problemsBecome grounded in RL fundamentals including Markov decision processes, dynamic programming, and temporal difference learningDive deep into a range of value and policy gradient methodsApply advanced RL solutions such as meta learning, hierarchical learning, multi-agent, and imitation learningUnderstand cutting-edge deep RL algorithms including Rainbow, PPO, TD3, SAC, and moreGet practical examples through the accompanying website
Used as a supplement for the reinforcement learning specialization on coursera and found it actually covers the same topics with the same approach as the online textbook for the series. Was a little disappointed by how derivative it is, as I was looking for a different perspective from 'Reinforcement Learning: An Introduction' by Sutton and Barto.
Not a good starting point to learn RL. You may consider watching some lectures or reading something more introductory to this field because the author seems to be targeting practitioners who have some background about RL. It contains introductory chapters, but I found them very brief and I relied on external resources to fill the gaps. Unfortunately, most of the source codes in the supplementary materials are outdated (in 2024) so you need to find something updated on github to get a hands-on experience. Overall, it's a nice read, especially the last 4 chapters that are about the real-life RL development lifecycle.
This is one of the most well-written and engaging technical books I’ve had the pleasure to read. What an exciting subject too. Reinforcement Learning - similar to Machine Learning a few years ago - is starting to see a big uptick in popular applications (see AlphaFold, end-to-end driverless cars). I recommend this book to anyone who’s remotely interested in AI. I think we’re going to see a lot more reinforcement learning in the next 5-10 years.