Discovering Strategic Behaviour of Multi-Agent Systems in Adversary Settings

Authors

  • Violeta Mirchevska Result d.o.o., Celovška cesta 182, 1000 Ljubljana & Jožef Stefan International Postgraduate School, Jamova 39, 1000 Ljubljana
  • Mitja Luštrek Jožef Stefan Institute, Jamova 39, 1000 Ljubljana
  • Andraž Bežek Jožef Stefan Institute, Jamova 39, 1000 Ljubljana & Marg d.o.o., Tržaška cesta 515, 1351 Brezovica pri Ljubljani
  • Matjaž Gams Jožef Stefan Institute, Jamova 39, 1000 Ljubljana & Jožef Stefan International Postgraduate School, Jamova 39, 1000 Ljubljana

Keywords:

Agent modelling, strategy discovery, behaviour analysis, multi-agent system, RoboCup

Abstract

Can specific behaviour strategies be induced from low-level observations of two adversary groups of agents with limited domain knowledge? This paper presents a domain-independent Multi-Agent Strategy Discovering Algorithm (MASDA), which discovers strategic behaviour patterns of a group of agents under the described conditions. The algorithm represents the observed multi-agent activity as a graph, where graph connections correspond to performed actions and graph nodes correspond to environment states at action starts. Based on such data representation, the algorithm applies hierarchical clustering and rule induction to extract and describe strategic behaviour. The discovered strategic behaviour is represented visually as graph paths and symbolically as rules. MASDA was evaluated on RoboCup. Both soccer experts and quantitative evaluation confirmed the relevance of the discovered behaviour patterns.

Downloads

Download data is not yet available.

Downloads

Published

2014-06-02

How to Cite

Mirchevska, V., Luštrek, M., Bežek, A., & Gams, M. (2014). Discovering Strategic Behaviour of Multi-Agent Systems in Adversary Settings. Computing and Informatics, 33(1), 79–108. Retrieved from http://147.213.75.17/ojs/index.php/cai/article/view/859

Most read articles by the same author(s)