Applying Diffusion to a Cooperative Coevolutionary Model

Abstract. Perhaps one the newest and of the more interesting cooperative approaches to evolutionary computation which has been more recently explored is the area of mutualism. In mutualistic methods, the problem is subdivided into modular components where each component evolves separately, but is evaluated in terms of the other components. In this way the problem may be cooperatively solved. In an attempt to give this novel approach more adaptive ability, in this paper we explore the effects of adding varying degrees of population diffusion via a mutatable tagging scheme applied to individual chromosomes.

• Type: Conference
• Year: 1998
• Where: PPSN
• Authors: R. Paul Wiegand
• Keywords: Distributed And Parallel Computing, Evolutionary Computation, Interesting Cooperative Approaches, Mutatable Tagging Scheme, PPSN 1998