2026-04-16T20:16:31.649Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:3:11501 2025-07-31T23:54:26Z 3

Predictive Modeling of Customer Churn: A Comparative Analysis of Ensemble Learning Techniques, Chapter in LNNS Lecture Notes in Networks and Systems: ADCIS 2024: Advances in Data-driven Computing and Intelligent Systems, Springer, volume 1304 2025 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/3/11501 https://link.springer.com/chapter/10.1007/978-981-96-3652-5_22 V. Thomas M. Zivkovic L. Jovanovic M. Antonijevic J. Perisic N. Bacanin J. Cadjenovic In this manuscript, an extensive analysis of customer churn forecasting utilizing ensemble learning methods is presented. This study aims to develop an effective model for identifying customers at risk of churn in the banking sector. The authors leverage a dataset comprising various customer attributes and behaviors to train and evaluate our predictive models. Ensemble learning algorithms were employed, namely XGBoost and AdaBoost, to build robust classifiers capable of accurately predicting customer churn. Through extensive experimentation and evaluation, the efficacy of the suggested approach regarding the predictive performance and model interpretability is demonstrated. Simulation findings reveal important insights into customer churn dynamics and provide valuable guidance for banking institutions to proactively manage customer relationships and reduce churn rates. bookPart Springer, Singapore 299 314 10.1007/978-981-96-3652-5_22 LNNS Lecture Notes in Networks and Systems: ADCIS 2024: Proceedings of International Conference on Advances in Data-driven Computing and Intelligent Systems, volume 1304