On the differential evolution schemes in MOEA/D

Recently, MOEA/D has gained some success in tackling a class of continuous multi-objective optimization problems with complicated pareto sets. It uses a simple differential evolution scheme (i.e., DE/BEST/1) to produce offspring solutions. However, the performance of MOEA/D is very sensitive to the...

Full description

Saved in:
Bibliographic Details
Published in2010 Sixth International Conference on Natural Computation Vol. 6; pp. 2788 - 2792
Main Authors Waibin Huang, Hui Li
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.08.2010
Subjects
Convergence
Educational institutions
Evolutionary computation
IEEE Press
Optical fibers
Optimization
Shape
Online AccessGet full text

Cover

More Information
Summary:Recently, MOEA/D has gained some success in tackling a class of continuous multi-objective optimization problems with complicated pareto sets. It uses a simple differential evolution scheme (i.e., DE/BEST/1) to produce offspring solutions. However, the performance of MOEA/D is very sensitive to the parameters in differential evolution schemes, such as scaling factor. In this paper, we further investigate the influence of various DE schemes in MOEA/D. Our experimental results show that the performance of MOEA/D relies on the choice of mating parents in DE schemes.
ISBN:1424459583
9781424459582
ISSN:2157-9555
DOI:10.1109/ICNC.2010.5583335