作者: Saul Zapotecas-Martinez , Hernan E. Aguirre , Kiyoshi Tanaka , Carlos A. Coello Coello
关键词: Algorithm 、 Usability 、 Flexibility (engineering) 、 Multi-objective optimization 、 Evolutionary algorithm 、 Evolutionary computation 、 Process (computing) 、 Decomposition (computer science) 、 Computational complexity theory 、 Mathematical optimization 、 Mathematics
摘要: In spite of the success multi-objective evolutionary algorithm based on decomposition (MOEA/D), generation weights for problems having many objectives, continues to be an open research problem. this paper, we introduce a new methodology low-discrepancy sequences generate vectors employed by MOEA/D. We analyze and compare proposed using different its impact in search process The approach is evaluated objective functions (up 15 objectives). show flexibility ease use type when adopting them