Conference Proceedings

A Clustering-Prediction Pipeline for Customer Churn Analysis

Hanming Zheng, Ling Luo, Goce Ristanoski, H Qiu (ed.), C Zhang (ed.), Z Fei (ed.), M Qiu (ed.), SY Kung (ed.)

KNOWLEDGE SCIENCE, ENGINEERING AND MANAGEMENT, PT III | SPRINGER INTERNATIONAL PUBLISHING AG | Published : 2021

Abstract

Customer churn is the event when customers using the products or services of an organization decide to no longer do so by either switching to another organization or by stopping using those products/services. It takes much more effort to attract new customers than to retain the existing clientele, which makes predicting customer churn and imposing retention strategies an essential task for many organizations. Though methodologies that deliver accurate predictions are available, there exists no general and transparent pipeline designed for churn prediction on different domains with high interpretability. In our work, we propose two domain-agnostic, inference-friendly, and easy-interpreting no..

View full abstract