Representation and Computation With Decision - and Game - Theoretic Agents

摘要

This project aimed at working towards a theory of representation and computation in decision and game-theoretic models. In many domains, such as the ones faced by the military, or in e-commerce, task allocation (and other multiagent protocols) is carried out in non-cooperative environments. Game theory deals with decision making in noncooperative environments, but ignores fault-tolerance and related computational issues. We have isolated several central and complementary lines of research, which are essential for establishing a general synthesized theory of representation and computation in decision and game-theoretic models, and made significant progress in each of those directions.