2009 | OriginalPaper | Buchkapitel
Agent Strategy Generation by Rule Induction in Predator-Prey Problem
verfasst von : Bartłomiej Śnieżyński
Erschienen in: Computational Science – ICCS 2009
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
This paper contains a proposal of application of rule induction for generating agent strategy. This method of learning is tested on a predator-prey domain, in which predator agents learn how to capture preys. We assume that proposed learning mechanism will be beneficial in all domains, in which agents can determine direct results of their actions. Experimental results show that the learning process is fast. Multi-agent communication aspect is also taken into account. We can show that in specific conditions transferring learned rules gives profits to the learning agents.