A consistency-based validation for data clustering

摘要: Clustering analysis is a powerful tool in customer segmentation. Although various algorithms have been proposed, the determination of optimal number clusters remains to be difficult issue. In this paper, clustering method based on consistency criterion proposed address The main characteristic new approach that it requires little prior information and can find automatically. Extensive comparisons are done over 22 real-world datasets from different domains, which four well-known combination with six indices used as benchmark methods. results demonstrate superiority our appropriately determining clusters. An application segmentation credit card users also illustrated.