Book summary

My congratulations on reaching the end of the book! I hope that the book was useful and you enjoyed reading it as much as I enjoyed gathering material and writing all the chapters. As a final word, I'd like to wish you good luck in this exciting and dynamic area of RL. The domain is developing very rapidly, but with an understanding of the basics, it becomes much simpler for you to keep track of the new developments and research in this field.

There are lots of very interesting topics left uncovered, such as partially observable MDPs (where environment observations don't fulfill the Markov property) or recent approaches to exploration, such as the count-based methods. There is a lot of recent activity around multi-agent methods, where many agents need to learn how to coordinate to solve a common problem. We also haven't mentioned the memory-based RL approach, where your agent can maintain some sort of a memory to keep its knowledge and experience...