Multi-objective evolutionary search strategies in constraint programming

Abstract It has been shown that evolutionary algorithms are able to construct suitable search strategies for classes of Constraint Satisfaction Problems (CSPs) in Programming. This paper is an explanation the use multi-objective optimisation contrast simple additive weighting techniques with a view develop CSPs. A hierarchical scheme employed select candidate strategy from Pareto frontier final evaluation. The results demonstrate significantly outperforms single objective same number evaluations. In situations where developed class problems fail extend unseen problem instances class, it found structure underlying CSPs do not resemble those training process.