作者: A. E. Eiben , J. I. van Hemert , E. Marchiori , A. G. Steenbeek
DOI: 10.1007/BFB0056863
关键词:
摘要: This paper presents a comparative study of Evolutionary Algorithms (EAs) for Constraint Satisfaction Problems (CSPs). We focus on EAs where fitness is based penalization constraint violations and the penalties are adapted during execution. Three different this approach implemented. For highly connected networks, results provide further empirical support to theoretical prediction phase transition in binary CSPs.