Title: A new automatic GEP-Cluster algorithm

Authors: Xin Du; Youcong Ni; Peng Ye; Ruliang Xiao

Addresses: State Key Laboratory of Software Engineering, Wuhan University, Wuhan, Hubei, China; Faculty of Software, Fujian Normal University, Fuzhou, Fujian, China ' Faculty of Software, Fujian Normal University, Fuzhou, Fujian, China ' College of Mathematics and Computer, Wuhan Textile University, Wuhan, Hubei, China ' Faculty of Software, Fujian Normal University, Fuzhou, Fujian, China

Abstract: GEP-Cluster, a clustering algorithm based on Gene Expression Programming (GEP), is a kind of automatic cluster algorithm for the clustering problem with unknown clustering number. However, its performance, especially the step for computing the distance between the data object and the corresponding cluster centre, can be improved. Parallel is undoubtedly a good method for improving the performance of algorithm. This paper proposes an improved auto-clustering algorithm based on Compute Unified Device Architecture (CUDA) and GEP, called ICGEP-Cluster. Experimental results show that ICGEP-Clustering has a better performance than GEP-Cluster, especially for the large scale of data objects.

Keywords: clustering algorithms; GEP; gene expression programming; CUDA; compute unified device architecture; ICGEP clustering.

DOI: 10.1504/IJWMC.2015.073100

International Journal of Wireless and Mobile Computing, 2015 Vol.9 No.3, pp.224 - 230

Received: 04 May 2015
Accepted: 01 Jun 2015
Published online: 19 Nov 2015 *

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