Learning in multi-agent systems

There is an increased interest in multi-agent systems (MASs) for comput-ing robust solutions to complex real world problems. In this paper we analyze dierent aspects of multi-agent systems, in particular multi-agent architec-tures, multi-agent problems, and optimization algorithms for MASs. Further-more, we present a scheme for mapping multi-agent problems to architectures which can be used for solving them and a mapping from multi-agent prob-lem features to optimization algorithms. Finally, we review the solutions of previous work on many dierent multi-agent problems.