Customer Segmentation Based on a Novel Hierarchical Clustering Algorithm

Customer segmentation plays an important role in customer relationship management(CRM). It allows companies to design and establish different strategies to maximize the value of customers. As one of the most important techniques of data mining, clustering analysis becomes wildly used method in customer segmentation. A novel customer segmentation method called fuzzy Fisher criterion based hierarchical clustering algorithm(FFCHC) is proposed. It applies fuzzy Fisher criterion algorithm (FFC) by successive dichotomy method and uses the clustering validity function to find out the optimal number of clusters. FFCHC can identify successive linear separable shapes effectively. The experimental results on a stock exchange customer dataset demonstrate its roles and performances on customer segmentation.