Integration of self-organizing feature maps and genetic-algorithm-based clustering method for market segmentation

Clustering analysis has been widely applied in the area of market segmentation. Conventional research usually uses the multivariate analysis method. Recently, due to promising results of computational intelligence techniques in engineering, they are also considered for market segmentation. Among them, genetic algorithms (GAs) are theoretically and empirically found to provide globally near-optimal solutions for various complex optimization problems. Because GA is good at searching, it can cluster the data according to their similarities. In addition, artificial neural networks also have high performance in both engineering and management. Hence, this research proposes a novel 2-stage method, which first uses self-organizing feature maps (SCIMs) to determine the number of clusters and then employs a GA-based clustering method to find the final solution (it is defined as S + G in this research). The results of simulated data via a Monte Carlo study show that the proposed method outperforms the other 2 methods: K means, which uses SOM to determine the number of clusters, and SOM followed by K means, based on both within-cluster variations (SSW) and the number of misclassifications. To further verify the proposed approach, a real-world problem, wireless telecommunications industry market segmentation, is employed. The results also show that the proposed method has the lowest SSW among the 3 methods.