Abstract: This thesis collects several papers that investigate multi-agent systems from different angles, using logics that combine concepts from game theory with well established modal treatment of time, knowledge and belief. Part I of the thesis represents the specification angle: we look for a language that allows to capture strategic abilities of agents under imperfect information in the most intuitive (and general) way. Part II is focused on the modeling dimension: we incorporate economic rationality in our view of agents, and study its impact on agents’ behavior within the framework of modal logic. The perspective of Part III is computational: we study the complexity of verification for several specification languages, and observe some intriguing patterns among them. Finally, Part IV presents an attempt at integration of two fundamental approaches to reasoning about systems: we show how certain aspects of qualitative (logical) reasoning can be incorporated into quantitative analysis of MAS in the style of Markov decision processes.
The thesis is not available for download due to copyright reasons, but if you feel an untamable urge to get hold of it, please contact me.
Individual and Collective Reasoning Group @ University of Luxembourg | |
Computational Intelligence Group @ Clausthal University of Technology | Last modified 2011-11-08 |