K-Means and Decision Trees in Action:Unveiling Customer Segments and Personalized Promotions for Credit Card Companies

Abstract

This study aims to analyze customer segmentation of a credit card company. The analysis consists of two phases: K-means clustering and Decision tree classification. In the initial phase, various numbers of clusters were explored using the elbow method, silhouette analysis, and Davies-Bouldin index. The second phase focused on Decision tree classification to identify key features that differentiate customer groups and capture characteristics of each cluster. This provided a clear understanding of the factors influencing customer behavior and preferences within each cluster. Finally, guidelines for developing customized campaigns or promotions were provided.