Customer Segmentation and Churn Prediction in Telecommunication Using Machine Learning

Kammili, Susanth (2024) Customer Segmentation and Churn Prediction in Telecommunication Using Machine Learning. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (553kB) | Preview
Preview	PDF (Configuration Manual) Download (142kB) | Preview

Abstract

Customer churn is a significant challenge for businesses, particularly in the telecommunications sector, where customer retention plays a crucial role in profitability and growth. This study addresses the problem of predicting customer churn using machine learning models, with an emphasis on both accuracy and interpretability. The research aims to develop and evaluate four machine learning models—logistic regression, random forests, gradient boosting, and LSTM networks—to predict customer churn. Further, the present research incorporates explainability tools like SHAP and LIME that would help in identifying what the algorithms look as the cause of churn to make the models efficient as well as to help the business executives understand them. The results show that the gradient boosting delivers the highest accuracy of 0.83% and the AUC-ROC of 0.90%; nevertheless, random forest and logistic regression models are also important and precise. While LSTM networks are quite effective, they failed because of the time agnostic data set used in the study. The implementation of SHAP and LIME also indicated that “Tenure,” and “Monthly Charges” were important features to distinguish churn. In this research, the following advancements were made 1) Moving from binary PAM and model performance metrics to show an integrated solution with an interval of interpretability and performance. The study also points to the lack of dynamic and temporal data as the way forward in improving the performance of churn models in practice.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Cosgrave, Noel UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HF Commerce > Marketing > Consumer Behaviour Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Telecommunications Industry
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	02 Sep 2025 15:17
Last Modified:	02 Sep 2025 15:17
URI:	https://norma.ncirl.ie/id/eprint/8721

Actions (login required)

View Item