Title: Estimation of distribution algorithms based on two copula selection methods

Authors: Xiaoping Wang; Huimin Gao; Jianchao Zeng

Addresses: College of Electrical and Information Engineering, Lanzhou University of Technology, Lanzhou, 730050, China; School of Business, Jiaxing University, Jiaxing, Zhejing, 314001, China ' Mechanical and Electrical Engineering College, Jiaxing University, Jiaxing, Zhejing, 314001, China ' Complex System and Computational Intelligence Lab, Taiyuan University of Science and Technology, Taiyuan, 030024, China

Abstract: In order to solve the copula selection problem, an innovative method, called switched copula selection method is introduced, which is inspired by the mixed copula selection method. The new method can ignore how to determine a fixed copula beforehand, whereas it organises different kinds of copulas through a copula selection pool. Applying copula selection methods to estimation of distribution algorithms (EDAs), the procedures of proposed mixed copula-EDAs and switched copula-EDAs are described in detail. Comparing with the single copula-EDAs, the experiment results show that switched copula-EDAs is easy to implement and provides a balance between accuracy and efficiency of the copula selection, in the meantime it promotes the performance whereas the computation complexity is almost the same.

Keywords: estimation of distribution algorithms; EDAs; copula selection; switched copula; mixed copula; optimisation.

DOI: 10.1504/IJCSM.2012.051616

International Journal of Computing Science and Mathematics, 2012 Vol.3 No.4, pp.317 - 331

Received: 24 Aug 2012
Accepted: 05 Sep 2012
Published online: 30 Aug 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article