Self-adaptive genetic algorithms with simulated binary crossover

Authors
Citation
K. Deb et Hg. Beyer, Self-adaptive genetic algorithms with simulated binary crossover, EVOL COMPUT, 9(2), 2001, pp. 197-221
Citations number
33
Categorie Soggetti
Computer Science & Engineering
Journal title
EVOLUTIONARY COMPUTATION
ISSN journal
10636560 → ACNP
Volume
9
Issue
2
Year of publication
2001
Pages
197 - 221
Database
ISI
SICI code
1063-6560(200122)9:2<197:SGAWSB>2.0.ZU;2-8
Abstract
Self-adaptation is an essential feature of natural evolution. However, in t he context of function optimization, self-adaptation features of evolutiona ry search algorithms have been explored mainly with evolution strategy (ES) and evolutionary programming (EP). In this paper, we demonstrate the self- adaptive feature of real-parameter genetic algorithms (GAs) using a simulat ed binary crossover (SBX) operator and without any mutation operator. The c onnection between the working of self-adaptive ESs and real-parameter GAs w ith the SBX operator is also discussed. Thereafter, the self-adaptive behav ior of real-parameter GAs is demonstrated on a number of test problems comm only used in the ES literature. The remarkable similarity in the working pr inciple of real-parameter GAs and self-adaptive ESs shown in this study sug gests the need for emphasizing further studies on self-adaptive GAs.