Evolutionary algorithms (EAs) for solving constraint satisfactionproblems (CSPs) can be roughly divided into two classes: EAs withadaptive fitness functions and heuristic-based EAs. A.E. Eiben et al.(1998) compared effective EAs of the first class experimentally using alarge set of benchmark instances consisting of randomly-generated binaryCSPs. In this paper, we complete this comparison by performing the sameexperiments using three of the most effective heuristic-based EAs. Theresults of our experiments indicate that the three heuristic-based EAshave similar performances on random binary CSPs. Comparing these resultswith those of A.E. Eiben et al., we are able to identify the best EA forbinary CSPs as the algorithm introduced by G. Dozier et al. (1994),which uses a heuristic as well as an adaptive fitness function
展开▼
