An empirical analysis of collaboration methods in cooperative coevolutionary algorithms

Abstract

Although a variety of coevolutionary methods have been explored over the years, it has only been recently that a general architecture for cooperative coevolution has been proposed. Since that time, the flexibility and success of this cooperative coevolutionary architecture (CCA) has been shown in an array of different kinds of problems. However, many questions about the dynamics of this model, as well as the efficacy of various CCA-specific choices remain unanswered. One such choice surrounds the issue of how the algorithm selects collaborators for evaluation. This paper offers an empirical analysis of various types of collaboration mechanisms and presents some basic advice about how to choose a mechanism which is appropriate for a particular problem. 1