Citation

Yee, Lim Jing (2025) Churn Prediction Using Machine Learning in Telecommunication Industry. In: The Smart Life Revolution. CRC Press, pp. 192-218. ISBN 978-104036402-4, 978-103283405-4

Full text not available from this repository.

Abstract

Customer churn is one of the causes that bring severe consequences to the financial losses of a business. The rapid growth of competitive industries, especially telecommunications, has made it harder to identify high-risk churners, and this has triggered the loss of business revenue in most companies. Therefore, customer churn prediction has earned the rising interest of researchers and business specialists because of its attractive ability to manage stable business profits. Although traditional machine learning (ML) techniques, such as Decision Trees, Random Forest, Logistic Regression and XGBoost, have successfully shown their ability to predict churners in past case scenarios, they still have many limitations when dealing with complex customer data. However, the growth of deep learning (DL) techniques, such as Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM), poses some potential in overcoming the weaknesses of traditional methods when it comes to handling dynamic and complicated data. This study provides a valuable overview of customer churn prediction, focusing on ML and DL approaches and comparing them with some traditional machine learning techniques used in previous studies. The overall flow of churn prediction using deep learning will be presented with two architectures of DL models, CNN and LSTM together with commonly used telecom churn datasets, and performance metrics. Lastly, the study will conclude with future directions suggested by researchers in prior studies.

Item Type:	Book Section
Uncontrolled Keywords:	Machine learning
Subjects:	Q Science > Q Science (General) > Q300-390 Cybernetics
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	26 Jun 2025 08:02
Last Modified:	26 Jun 2025 08:02
URII:	http://shdl.mmu.edu.my/id/eprint/14122

Downloads

Edit (login required)