A Hybrid Machine Learning Approach for Customer Segmentation Using RFM Analysis

ABSTRACT

Due to COVID-19 situation, online retailing (electronic retailing) for purchasing goods has recently increased which leads to the need of customer segmentation. Customer segmentation is done based on customers’ past purchase behavior and then divide them into different categories, i.e., loyal customer, potential customer, new customer, customer needs attention, customers require activation. This paper uses recency, frequency, monetary value (RFM) analysis and K-means clustering technique for grouping the customers. Further to enhance the efficiency of segmentation, a decision tree is used to create nested splitting (based on Gini index) inside the each cluster. The implementation of proposed hybrid approach is showing promising results for customer segmentation to take better management decisions. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.