A Historical Population in a Coevolutionary System

Author

Avery, Phillipa M. ; Michalewicz, Zbigniew ; Schmidt, Martin

Author_Institution

Dept. of Comput. Sci., Adelaide Univ., SA

fYear

2007

fDate

1-5 April 2007

Firstpage

104

Lastpage

111

Abstract

The use of memory in coevolutionary systems is considered an important mechanism to counter the Red Queen effect. Our research involves incorporating a memory population that the coevolving populations compete against to obtain a fitness that is influenced by past generations. This long term fitness then allows the population to have continuous learning that awards individuals that do well against the current populations, as well as previous winning individuals. By allowing continued learning, the individuals in the populations increase their overall ability to play the game of TEMPO, not just to play a single round with the current opposition.

Keywords

evolutionary computation; game theory; learning (artificial intelligence); TEMPO game; adaptive memory; coevolutionary system; computational intelligence; continuous learning; historical population; memory population; winning individual; Australia; Computational intelligence; Computer science; Counting circuits; Fuzzy systems; History; Humans; Intelligent systems; Military computing; System testing; Adaptive memory; Computational Intelligence; TEMPO;

fLanguage

English

Publisher

ieee

Conference_Titel

Computational Intelligence and Games, 2007. CIG 2007. IEEE Symposium on

Conference_Location

Honolulu, HI

Print_ISBN

1-4244-0709-5

Type

conf

DOI

10.1109/CIG.2007.368085

Filename

4219030