Title: A bi-tour ant colony optimisation framework for vertical partitions
Author: Chun-Hung Cheng, Angappa Gunasekaran, Kwan-Ho Woo
Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, Shatin, NT, Hong Kong SAR, PR China.
Department of Decision and Information Sciences, University of Massachusetts Dartmouth, North Dartmouth, MA 02747, USA.
XML Asia Ltd., Room 1601, 16/F., Jubilee Centre, 18 Fenwick Street, Wan Chai, Hong Kong SAR, PR China
Abstract: Clustering refers to a process of grouping together similar objects while separating out the dissimilar objects. In this work, we consider block clustering in vertical partitioning. Block clustering is a specific clustering method, which clusters the sets of objects and their associated attributes (descriptors) together, simultaneously, in a solution matrix. For this specific problem we propose using a bi-tour ant colony optimisation. To show the quality of the new proposed approach, we conduct an extensive computational study and show that our method is performed better than some traditional clustering methods, such as genetic algorithms and average linkage clustering.
Keywords: block clustering; ACO; ant colony optimisation; vertical partitioning; bi-tour frameworks; vertical partitions; similar objects; dissimilar objects; clusters; associated attributes; descriptors; solution matrixes; computational studies; genetic algorithms; average linkages; industrial engineering; systems engineering.
Int. J. of Industrial and Systems Engineering, 2011 Vol.7, No.3, pp.341 - 356
Available online: 10 Mar 2011