•  
  •  
 

Applied Mathematics & Information Sciences

Author Country (or Countries)

ROC

Abstract

This study attempts to propose a novel algorithm, hybrid of artificial immune network (aiNet) and particle swarm K-means optimization (PSKO)(aiNet-PSKO) algorithm, for cluster analysis. In order to verify the proposed methods, four benchmark data sets, Iris, Glass, Wine, and Breast Cancer, are first employed. The computational results indicate that aiNet-PSKO algorithm outperforms artificial immune system and particle swarm optimization related algorithms. Thereafter, these methods are further applied to the transaction database for an internet florist. The results also show that the aiNet-PSKO algorithm has the lowest sum of Euclidean distance value. The results can be used to make the marketing strategies for different clusters in order to provide different products or services customers prefer.

Digital Object Identifier (DOI)

http://dx.doi.org/10.18576/amis/110328

Share

COinS