We present a study of cooperative coevolution applied to moderately complex optimization problems in large-population environments. The study asks three questions. First: what collaboration methods perform best, and when? Second: how many subpopulations are desirable? Third: is it worthwhile to do more than one trial per fitness evaluation? We discovered that parallel methods tended to work better than sequential ones, that "shuffling" (a collaboration method) predominated in performance in

more »

... complex problems, that more subpopulations generally did better, and that more trials performed marginally better.