Springer, Berlin Mathematical Foundation of Reinforcement Learning

This book provides a mathematical but friendly introduction to the fundamental concepts, basic problems, and classic algorithms in reinforcement learning. It can help readers understand the theoretical roots of an algorithm and hence why the algorithm was designed in the first place and why it works. Many illustrative examples are given, and the mathematics is presented in a carefully designed manner to ensure that the book is friendly to read.The contents of this book include two aspects. The first is about the mathematical foundation of reinforcement learning, which includes the Bellman equation, Bellman optimality equation, value iteration and policy iteration methods, and stochastic approximation. The second is about classic reinforcement learning algorithms, which include Monte Carlo methods, temporal-difference methods, function approximation, policy gradient, and actor-critic methods.Given its scope, the book will appeal to undergraduate and graduate students, post-doctoral researchers, lecturers, industrial researchers, and anyone interested in reinforcement learning.