Multiagent learning for open systems: a study in opponent classification

Authors:
Michael Rovatsos;Gerhard Weiß;Marco Wolf
Affiliations:
Department of Informatics, Technical University of Munich, Garching bei München, Germany;Department of Informatics, Technical University of Munich, Garching bei München, Germany;Department of Informatics, Technical University of Munich, Garching bei München, Germany
Venue:
Adaptive agents and multi-agent systems
Year:
2003

Citing 9
Cited 2

Open information systems semantics for distributed artificial intelligence

Artificial Intelligence
Social conceptions of knowledge and action: DAI foundations and open systems semantics

Artificial Intelligence
Technical Note: \cal Q-Learning

Machine Learning
Machine Learning

Machine Learning
An approach to the analysis and design of multiagent systems based on interaction frames

Proceedings of the first international joint conference on Autonomous agents and multiagent systems: part 2
Efficient algorithms for learning to play repeated games against computationally bounded adversaries

FOCS '95 Proceedings of the 36th Annual Symposium on Foundations of Computer Science
Hierarchical Common-Sense Interaction Learning

ICMAS '00 Proceedings of the Fourth International Conference on MultiAgent Systems (ICMAS-2000)
Learning models of intelligent agents

AAAI'96 Proceedings of the thirteenth national conference on Artificial intelligence - Volume 1
Learning other agents' preferences in multiagent negotiation

AAAI'96 Proceedings of the thirteenth national conference on Artificial intelligence - Volume 1

An evidential and genetic method for cooperative learning systems

Multiagent and Grid Systems
On the organisation of agent experience: scaling up social cognition

Socionics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Open systems are becoming increasingly important in a variety of distributed, networked computer applications. Their characteristics, such as agent diversity, heterogeneity and fluctuation, confront multiagent learning with new challenges. This paper presents the interaction learning meta-architecture InFFrA as one possible answer to these challenges, and introduces the opponent classification heuristic ADHoc as a concrete multiagent learning method that has been designed on the basis of InFFrA. Extensive experimental validation proves the adequacy of ADHoc in a scenario of iterated multiagent games and underlines the usefulness of schemas such as InFFrA specifically tailored for open multiagent learning environments. At the same time, limitations in the performance of ADHoc suggest further improvements to the methods used here. Also, the results obtained from this study allow more general conclusions regarding the problems of learning in open systems to be drawn.