Intelligent customer retention prediction using distributed federated learning

Prakash Choudhary; Amandeep Pooni; Ravi Raj Choudhary; Prakash Choudhary; Amandeep Pooni; Ravi Raj Choudhary

doi:10.3934/InnoEcon.2026004

Innovation Economics

2026, Volume 1, Issue 1: 68-85. doi: 10.3934/InnoEcon.2026004

Previous Article Next Article

Research article

Intelligent customer retention prediction using distributed federated learning

1.
Department of Computer Science and Engineering, Central University of Rajasthan, Rajasthan, India
2.
Department of Computer Science and Engineering, National Institute of Technology, Hamirpur, Himachal Pradesh, India
3.
Department of Computer Science, Central University of Rajasthan, Rajasthan, India

Received: 07 April 2026 Revised: 16 June 2026 Accepted: 22 June 2026 Published: 14 July 2026
JEL Codes: C45, C53, C81, C84, C88, L84

Customer retention is a key business objective, as retaining existing customers improves profitability and long-term growth. Predicting customer churn enables organizations to identify at-risk customers and implement effective retention strategies. However, customer data are often distributed across organizations and cannot be shared due to privacy, regulatory, and competitive concerns.
This study explored privacy-preserving approaches for customer retention prediction using internet service provider (ISP) data. We applied federated learning and a distributed form of incremental machine learning on internet service provider data to study and compare privacy conserving methods of predicting customer churn. With the federated learning approach, we divided and distributed data across different clients, thereby training individual models on those clients and the resulting models were averaged on a central server. In the distributed incremental learning approach, instead of transferring all models to a central server, a single model is trained and passed on to the next server to train on new data utill all clients are covered and the resultant model is synchronized across all clients. The findings provide practical insights for organizations seeking to balance predictive performance, customer retention, and data privacy requirements.
- stacked ensemble model,
- user retention,
- federated learning,
- business analytics
Citation: Prakash Choudhary, Amandeep Pooni, Ravi Raj Choudhary. Intelligent customer retention prediction using distributed federated learning[J]. Innovation Economics, 2026, 1(1): 68-85. doi: 10.3934/InnoEcon.2026004

Related Papers:

Abstract

Customer retention is a key business objective, as retaining existing customers improves profitability and long-term growth. Predicting customer churn enables organizations to identify at-risk customers and implement effective retention strategies. However, customer data are often distributed across organizations and cannot be shared due to privacy, regulatory, and competitive concerns.

This study explored privacy-preserving approaches for customer retention prediction using internet service provider (ISP) data. We applied federated learning and a distributed form of incremental machine learning on internet service provider data to study and compare privacy conserving methods of predicting customer churn. With the federated learning approach, we divided and distributed data across different clients, thereby training individual models on those clients and the resulting models were averaged on a central server. In the distributed incremental learning approach, instead of transferring all models to a central server, a single model is trained and passed on to the next server to train on new data utill all clients are covered and the resultant model is synchronized across all clients. The findings provide practical insights for organizations seeking to balance predictive performance, customer retention, and data privacy requirements.

References

[1]	Ali N, Shabn OS (2024) Customer lifetime value (clv) insights for strategic marketing success and its impact on organizational financial performance. Cogent Bus Manag, 11. https://doi.org/10.1080/23311975.2024.2361321 doi: 10.1080/23311975.2024.2361321
[2]	Amin A, Adnan A, Anwar S (2023) An adaptive learning approach for customer churn prediction in the telecommunication industry using evolutionary computation and naive bayes. Appl Soft Comput, 137: 110103. https://doi.org/10.1016/j.asoc.2023.110103 doi: 10.1016/j.asoc.2023.110103
[3]	Chaubey G, Gavhane PR, Bisen D (2022) Customer purchasing behavior prediction using machine learning classification techniques. J Amb Intel Hum Comp 14: 16133–16157. https://doi.org/10.1007/s12652-022-03837-6 doi: 10.1007/s12652-022-03837-6
[4]	Do D, Huynh P, Vo P, et al. (2017) Customer churn prediction in an internet service provider. 2017 IEEE International Conference on Big Data (Big Data), Boston, MA, USA, 3928–3933. https://doi.org/10.1109/BigData.2017.8258400
[5]	Dou Q, So TY, Jiang M, et al. (2021) Federated deep learning for detecting covid-19 lung abnormalities in ct: A privacy preserving multinational validation study. npj Digit Med 4: 60. https://doi.org/10.1038/s41746-021-00431-6 doi: 10.1038/s41746-021-00431-6
[6]	He H, Chen S, Li K, et al. (2018) Incremental learning from stream data. Ieee T Neur Net Lear 22: 1901–1914. https://doi.org/10.1109/TNN.2011.2171713 doi: 10.1109/TNN.2011.2171713
[7]	Hu K, Li Y, Xia M, et al. (2021) Federated learning: A distributed shared machine learning method. Complexity 2021: 20. https://doi.org/10.1155/2021/8261663 doi: 10.1155/2021/8261663
[8]	Imani M, Joudaki M, Beikmohammadi A, et al. (2025) Customer churn prediction: A systematic review of recent advances, trends, and challenges in machine learning and deep learning. Mach Learn Knowl Extr, 7: 105. https://doi.org/10.3390/make7030105 doi: 10.3390/make7030105
[9]	Krishnan M (2024) Enhancing customer churn prediction in telecom using federated learning. Available from: https://norma.ncirl.ie/id/eprint/7953.
[10]	Kunt MS (2021) Internet service provider customer churn dataset. Available from: https://www.kaggle.com/datasets/mehmetsabrikunt/internet-service-churn.
[11]	Lalwani P, Mishra MK, Chadha JS, et al. (2022) Customer churn prediction system: A machine learning approach. Computing 104: 271–294. https://doi.org/10.1007/s00607-021-00908-y doi: 10.1007/s00607-021-00908-y
[12]	Ma C, Li J, Ding M, et al. (2020) On safeguarding privacy and security in the framework of federated learning. Ieee Network, 34: 242–248. https://doi.org/10.1109/MNET.001.1900506 doi: 10.1109/MNET.001.1900506
[13]	McMahan HB, et al. (2017). Communication-efficient learning of deep networks from decentralized data. In: Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS), Fort Lauderdale, Florida. https://doi.org/10.48550/arXiv.1602.05629
[14]	Naz NA, Shoaib U, Sarfraz MS (2018) A review on customer churn prediction data mining modeling techniques. Indian J Sci Technol 11: 1–7. https://doi.org/10.17485/ijst/2018/v11i27/121478 doi: 10.17485/ijst/2018/v11i27/121478
[15]	Okonkwo R, Englama KS (2025) Ai-enhanced real-time customer churn prediction via federated learning for privacy-preserving and optimized marketing decision. Eur J Comput Sci Inform Technol, 13: 32–39. https://doi.org/10.37745/ejcsit.2013 doi: 10.37745/ejcsit.2013
[16]	Pekar A, Makara LA, Biczok G (2024) Incremental federated learning for traffic flow classification in heterogeneous data scenarios. Neural Comput Applic, 36: 20401–20424. https://doi.org/10.1007/s00521-024-10281-4 doi: 10.1007/s00521-024-10281-4
[17]	Rodan A, Fayyoumi A, Faris H, et al. (2014) Negative correlation learning for customer churn prediction: A comparison study. Sci World J 7. https://doi.org/10.1155/2015/473283 doi: 10.1155/2015/473283
[18]	Sheller MJ, Edwards B, Reina GA, et al. (2020) Federated learning in medicine: Facilitating multi-institutional collaborations without sharing patient data. Sci Rep 10: 12598. https://doi.org/10.1038/s41598-020-69250-1 doi: 10.1038/s41598-020-69250-1
[19]	Shi N, Lai F, Al Kontar R, et al. (2021) Fed-ensemble: Improving generalization through model ensembling in federated learning. J Big Data 6: 28. https://doi.org/10.1186/s40537-019-0191-6 doi: 10.1186/s40537-019-0191-6
[20]	Sikri A, Jameel R, Idrees SM, et al. (2024) Enhancing customer retention in telecom industry with machine learning driven churn prediction. Sci Rep 14: 13097. https://doi.org/10.1038/s41598-024-63750-0 doi: 10.1038/s41598-024-63750-0
[21]	Wu Z, He T, Sun S, et al. (2024) Federated class-incremental learning with self-distillation (fedclass). arXiv. https://doi.org/10.48550/arXiv.2401.00622
[22]	Yang X, Feng Y, Fang W, et al. (2020) An accuracy-lossless perturbation method for defending privacy attacks in federated learning. WWW '22: Proceedings of the ACM Web Conference 2022, 732–742. https://doi.org/10.1145/3485447.3512233

Reader Comments

Your name:*

Email:*
© 2026 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)