2017/03/28 by Mansoureh Aghabeig, Aghabeig, Mansoureh, Andrzej Jaszkiewicz +1
Computer Science · #Advanced Multi-Objective Optimization Algorithms #Metaheuristic Optimization Algorithms Research #Evolutionary Algorithms and Applications
paper · pdf · doi:10.48550/arxiv.1703.09469
In this paper we systematically study the importance, i.e., the influence on\nperformance, of the main design elements that differentiate scalarizing\nfunctions-based multiobjective evolutionary algorithms (MOEAs). This class of\nMOEAs includes Multiobjecitve Genetic Local Search (MOGLS) and Multiobjective\nEvolutionary Algorithm Based on Decomposition (MOEA/D) and proved to be very\nsuccessful in multiple computational experiments and practical applications.\nThe two algorithms share the same common structure and differ only in two main\naspects. Using three different multiobjective combinatorial optimization\nproblems, i.e., the multiobjective symmetric traveling salesperson problem, the\ntraveling salesperson problem with profits, and the multiobjective set covering\nproblem, we show that the main differentiating design element is the mechanism\nfor parent selection, while the selection of weight vectors, either random or\nuniformly distributed, is practically negligible if the number of uniform\nweight vectors is sufficiently large.\n