Combinations of evolutionary algorithms and fuzzy systems: a survey

Abstract

Lot of schemes to combine evolutionary algorithms and fuzzy systems have been reported in the literature in recent years. Evolutionary algorithms have been used to design fuzzy systems and fuzzy systems have been utilized to adapt the evolutionary algorithms. Both sides interact with each other both supportively and collaboratively. We provide a survey with focus on evolutionary algorithms as supporting tools for designing fuzzy systems since this is the main research work reported in the literature. The survey is organized from three perspectives: homogeneous vs. heterogeneous representation; online learning vs. offline learning; static learning vs. adaptive learning algorithms.