Reinforcement Learning Resources
- Reinforcement Learning by Sutton and Bartow covers today's key algorithms and theory, illustrating these concepts using real-world applications that range from learning to control robots, to learning to defeat the human world-champion Go player, and discussing fundamental connections between these computer algorithms and research on human learning from psychology and neuroscience. Truly an amazing read.
- Maxim Lapan's book on "Deep Reinforcement Learning Hands-On" is written using the PyTorch framework, making it so accessible to beginners. PyTorch makes it easier to read and digest because of the cleaner code which simply flows allowing the reader to focus more on the logic of the algorithms rather than on the nuts and bolts of the code examples, which themselves are clearly written and rely on the OpenAI Gym framework to make replication straightforward.
Online Courses
- Univ of Alberta's Fundamentals of Reinforcement Learning offered by Coursera is a great start to your journey in RL. Taught by Prof. Martha White and Prof. Adam White, it has really nice explanations with reading references to Sutton's book.
- Udacity has a Deep Reinforcement Learning Nanodegree, which I feel is a natural next step to the Coursera course. This offering is remarkable in that, it strikes a good balance among the theoretical aspects, writing codes from scratch and using available libraries. (While Udacity is slightly pricey, there are always discounts going on which can be utlizised.)
- As is probably expected Stanford has a ton of materials on Reinforcement Learning; I like the Standord CS 234 course in particular, which has lots of cool projects and assignments. Also, it is advisable to check out David Silver's course at UC London. Vid Lecs available freely!
