Reinforcement Learning (RL) is a branch of Artificial Intelligence (AI) that teaches agents to learn optimal behavior through interaction, feedback, and long-term goals.
Reinforcement Learning (RL) is a branch of Artificial Intelligence (AI) where agents learn optimal behavior through interaction with an environment by receiving feedback in the form of reward. After decades of research, RL has matured into a powerful technology driving real-world innovation; it is now used in areas such as robotics, energy systems, finance, and autonomous vehicles.
Yet, for many, RL feels inaccessible, buried under dense mathematics and complex theory. This book changes that. It is designed to help newcomers start applying RL as quickly as possible through a classical pedagogical approach: many small, focused examples that build intuition and practical skill step by step.
Featuring:
- Essential concepts explained from the ground up
- Code-based examples that reveal how algorithms work in practice
- Worked examples by hand to strengthen intuition, just like in engineering or mathematics textbooks
- Language-agnostic guidance, easily followed using Python, Java, or C++
Even readers without coding or university-level mathematics backgrounds will gain valuable insight into the fascinating world of RL—insight that may become a critical differentiator in the age of AI. Whether you are a student or professional, Reinforcement Learning Explained will give you the tools and confidence to explore one of AI’s most exciting frontiers.
About the Authors. Introduction. Preface. Acknowledgements. Cover. 1
From Rules to Learning. 2 From Markov to Bellman. 3 Reinforcement Learning
Concepts. 4 Temporal Difference Learning. 5 Monte Carlo Methods. 6 n-Step
Learning. 7 Safe-Action Reinforcement Learning. 8 Non-Episodic Learning. 9
Next-Level Concepts. 10 Policy Gradient Methods. 11 Actor-Critic Methods. 12
Deep Reinforcement Learning. 13 Monte Carlo Tree Search. 14 Combining
Learning and Search. 15 Multi-Agent Reinforcement Learning. 16 Outlook.
Appendix. Index.
Jonas Hellgren is a researcher specializing in reinforcement learning, optimization, and electrified vehicle systems. With experience across academia and industry spanning patents, publications, thesis supervision, and industrial projects, he brings both practical insight and theoretical depth. This book reflects his commitment to making complex ideas accessible.
Johannes Lindgren is a technical consultant specializing in software development, verification, and commissioning across rail, automotive, and maritime applications. Currently at Combine, developing software for the rail sector. Previous roles include simulation and verification at Volvo Autonomous Solutions and system commissioning at Lean Marine, along with research in image segmentation at CPAC Systems.
