Learning and coordination. Enhancing agent performance through distributed decision making (Q1337086)

This book presents a general framework for adaptive systems, by examining a number of models exhibiting the stochastic nature of learning behavior. The coordinative strategies of intelligent systems acting/working in uncertain environments are explored. Chap. 1 is an Introduction and framework for adaptive systems, presenting a general algebraic model for learning systems, based on autonomous agents acting individually and collectively. Chap. 2, Learning speed in neural networks, presents a hybrid paradigm involving both connectionism and hierarchical structures. The need for descentralized coordination, the behavior of collaborative systems interpreted in terms of games of strategy, and the conditions for globally optimal behavior are presented in the Chap. 3, Principles of coordination. Chapter 4, Case Study in Coordination, examines a game theoretic deployment approach to robotic teams in an unstructured environment. In the final Chap. 5, Conclusion, the author summarizes his research work and briefly discusses its implications. An Appendix, Dynamic models in statistical physics, presents the theoretical basis for the statistical models discussed in Chap. 2. This fine book is published in the KLUWER Intl. Series on Microprocessor-Based and Intelligent Systems Engineering addressing a wide audience.