Title :
Evolving group strategies for IPD
Author :
Hingston, Philip
Author_Institution :
Sch. of Comput. Sci. & Security, Edith Cowan Univ., Perth, WA, Australia
Abstract :
The Iterated Prisoners Dilemma (IPD) is often used to model cooperation between self-interested agents. In an earlier study, we introduced a framework using IPD to study the effects of species-level competition on the evolution of cooperative behaviour. In this paper, we extend the previous work, using co-evolutionary simulations of interactions between species of IPD-playing agents to investigate how group strategies may evolve. We find that the ability to cooperate more with agents of the same species greatly increases the ferocity of competition between species.
Keywords :
competitive algorithms; evolution (biological); game theory; group theory; IPD playing agent; coevolutionary simulation; cooperative behaviour; group strategies; iterated prisoners dilemma; species level competition; Bioinformatics; Biological system modeling; Computational modeling; Evolution (biology); Games; Genomics; Organisms;
Conference_Titel :
Evolutionary Computation (CEC), 2010 IEEE Congress on
Conference_Location :
Barcelona
Print_ISBN :
978-1-4244-6909-3
DOI :
10.1109/CEC.2010.5586197