A population ecology inspired parent selection strategy for numerical constrained optimization problems

There are no files associated with this record.

Title A population ecology inspired parent selection strategy for numerical constrained optimization problems
Author Yuchi, Ming; Kim, Jong-Hwan; Jo, Jun Hyung
Journal Name Applied Mathematics and Computation (AMC)
Editor John L. Casti, Melvin Scott
Year Published 2007
Place of publication Amsterdam
Publisher Elsevier
Abstract A population ecology inspired parent selection strategy is proposed to improve the searching ability of evolutionary algorithms for numerical constrained optimization problems. This method is mainly used to help find an appropriate number of feasible parents for offspring generation. Based on the similar phenomenon in population ecology, the number of feasible parents has a sigmoid-type relationship with that of the feasible individuals. To implement the novel parent selection strategy, the population is divided into two groups according to the feasibility of the individuals: the feasible group and infeasible group. The evaluation and ranking of these two groups are performed separately. The dynamic penalty method, annealing penalty method and stochastic ranking method are tested with the parent selection strategy on 13 benchmark problems. The results show that the proposed method is capable of improving the searching performance.
Peer Reviewed Yes
Published Yes
Volume 190
Issue Number 1
Page from 292
Page to 304
ISSN 0096-3003
Date Accessioned 2008-03-28
Language en_AU
Research Centre Institute for Integrated and Intelligent Systems
Faculty Faculty of Science, Environment, Engineering and Technology
Subject Neural Networks, Genetic Alogrithms and Fuzzy Logic
URI http://hdl.handle.net/10072/18503
Publication Type Journal Articles (Refereed Article)
Publication Type Code c1

Show simple item record

Griffith University copyright notice