Competitive Markov decision processes

Stochastic Games have been studied by mathematicians, operations researchers, electrical engineers, and economists since the 1950s; the simpler single-controller, noncompetitive version of these models evolved separately under the name of Markov Decision Processes. This book is devoted to a unified treatment of both subjects under the general heading of Competitive Markov Decision Processes. It examines these processes from the standpoints of modeling and of optimization, providing newcomers to the field with an accessible account of algorithms, theory, and applications, while also supplying specialists with a comprehensive survey of recent developments. Requiring only some knowledge of linear algebra and real analysis (further mathematical details are supplied in appendices), and limiting itself to finite-state discrete-time models, the book is suitable as a graduate text. Some of the more advanced topics may also be omitted without affecting the continuity of the presentation, making the text accessible to advanced undergraduates.