The task of managing general game playing in a multi-agent system is the problem addressed in this paper. It is considered to be done by an agent. There are many reasons for constructing such an agent, called general game management agent. This agent manages strategic interactions between other

This paper presents a methodology for agent-oriented analysis and design. The methodology is general, in that it is applicable to a wide range of multi-agent systems, and comprehensive, in that it deals with both the macro-level (societal) and the micro-level (agent) aspects of systems

. The model is initialized in an optimistic fashion: all actions in all states return the maximal possible reward (hence the name). During execution, it is updated based on the agent's observations. R-max improves upon several previous algorithms: (1) It is simpler and more general than Kearns and Singh

Abstract. Many autonomous agents operate in domains in which the cooperation of their fellow agents cannot be guaranteed. In such domains negotiation is essential to persuade others of the value of co-operation. This paper describes a general framework for negotiation in which agents exchange

The traditional representations of games using the extensive form or the strategic (normal) form obscure much of the structure that is present in real-world games. In this paper, we propose a new representation language for general multiplayer games — multi-agent influence diagrams (MAIDs

Market mechanisms, such as auctions, will likely represent a common interaction medium for agents on the Internet. The Michigan Internet AuctionBot is a flexible, scalable, and robust auction server that supports both software and human agents. The server manages many simultaneous auctions

We extend Q-learning to a noncooperative multiagent context, using the framework of generalsum stochastic games. A learning agent maintains Q-functions over joint actions, and performs updates based on assuming Nash equilibrium behavior over the current Q-values. This learning protocol provably

This paper describes an approach to rein-forcement learning in multiagent general-sum games in which a learner is told to treat each other agent as either a \friend " or \foe". This Q-learning-style algorithm provides strong convergence guarantees compared to an ex-isting Nash

rate. We examine this technique theoretically, proving convergence in self-play on a restricted class of iterated matrix games. We also present empirical results on a variety of more general stochastic games, in situations of self-play and otherwise, demonstrating the wide applicability of this method.

Game theory is one of the key paradigms behind many scientific disciplines from biology to behavioral sciences to economics. In its evolutionary form and especially when the interacting agents are linked in a specific social network the underlying solution concepts and methods are very similar