A Swarm Optimizer Based on Multi-Criterion Decision Making, part I: Theoretical investigation

This paper introduces a new swarm optimizer based on multi-criterion decision making (MCDM). Agents inside the swarm are making decision toward their candidate moves in the search space. Decision making alternatives are defined from set of swarm basis behaviors. Fitness and diversity are two criteria that used to measure the performance of every alternative. Levenshtien edit distance is used to compute the distance between two positions in genotype space. Criteria are then standardized by means of fuzzy sets. Fuzzyfied values of criteria are then aggregated by the fuzzy ordered weighted average (OWA) to reach single evaluation function. Overall decision making process is made to promote both fitness and diversity