Model-Based Reinforcement Learning - From Data to Continuous Actions with a Python-based Toolbox

Model-Based Reinforcement Learning is a comprehensive and practical approach to reinforcement learning that provides a model-based framework to bridge the topics of systems identification, model-based reinforcement learning, and optimal control. It offers a end-to-end framework of a more tractable model-based reinforcement learning technique, with detailed comparisons of different techniques, applications, and case studies. It is a useful reference for senior undergraduate students, graduate students, research assistants, professors, process control engineers, and roboticists.

Format: Hardback
Length: 272 pages
Publication date: 09 December 2022
Publisher: John Wiley and Sons Ltd

Reinforcement learning is a crucial machine learning paradigm that enables intelligent agents to optimize device behavior. While it has achieved remarkable success and popularity, previous research has primarily focused on theory or algorithms, with a limited emphasis on practical applications. To address this gap, Model-Based Reinforcement Learning offers a comprehensive and practical approach to the field. By providing a model-based framework, the book aims to bridge the gap between optimal control and dynamic programming theory and simulation-based algorithms.

The authors seek to develop a model-based framework for data-driven control that encompasses systems identification from data, model-based reinforcement learning, and optimal control. This interdisciplinary approach aims to enhance the efficiency and effectiveness of reinforcement learning systems.

The book is designed to serve as a valuable textbook for graduate courses in data-driven and learning-based control, emphasizing the modeling and control of dynamical systems from data. It offers detailed comparisons of various techniques, including basic linear quadratic controller, learning-based model predictive control, model-free reinforcement learning, and structured online learning. Additionally, the book includes applications and case studies on ground vehicles with nonholonomic dynamics and another on quadrator helicopters.

To facilitate learning, the book is accompanied by an online platform, Pyt, which provides additional resources, examples, and exercises. Overall, Model-Based Reinforcement Learning offers a comprehensive and practical guide to the field, enabling researchers and practitioners to develop more efficient and effective reinforcement learning systems.

Weight: 467g
Dimension: 229 x 152 x 16 (mm)
ISBN-13: 9781119808572