Reinforcement Learning

Reinforcement Learning: An Introduction, 2nd Edition, is an artificial intelligence textbook published by The MIT Press. Written by Richard S. Sutton and Andrew G. Barto, this volume provides an expanded and updated revision of their foundational work.

The text frames reinforcement learning as a computational approach where an agent learns through interaction with an environment. It examines how the agent navigates complex and uncertain surroundings to maximize cumulative reward.

Included in the Adaptive Computation and Machine Learning series, the 552-page textbook features 64 color illustrations and 51 black-and-white figures to support reader study.