Selecting Heterogeneous Team Players by Case-Based Reasoning: A Case Study in Robotic Soccer Simulation

It is a vital behaviour pattern of humans, the most highly developed autonomous agents, to make use of experiences accumulated in the past and to solve new problems in analogy to solutions of old, yet similar problems. This report gives an outline of our work to apply that case-based approach to an arti cial agent in the domain of Robotic Soccer simulation. We enable the online coach of a robotic soccer team to determine the team line-up by a technique that incorporates knowledge into its reasoning process that was gained from former soccer matches. In order to use the knowledge contained in old cases, it is indispensable to de ne a meaningful evaluation of old solutions. Moreover, it is necessary to retrieve and adapt those solutions whose application to the current problem situation promises to be most auspicious. For these reasons, we also concentrate on the assessment of a team's performance. Further, we focus on a most precise calculation of the similarity between heterogeneous player types. This research was performed while Gabel visited Carnegie Mellon from the University of Kaiserslautern, Germany. Veloso's research was sponsored by the United States Air Force under Cooperative Agreements Nos. F30602-00-2-0549 and F3060298-2-0135. The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the oAEcial policies or endorsements, either expressed or implied, of the funding sources.