Citation

G Murugasu, Umapathy Sivan and Subbarao, Anusuyah (2026) Predicting Customer Churn through User’s Digital Demography Using Machine Learning. JOIV : International Journal on Informatics Visualization, 10 (4). p. 1792. ISSN 2549-9610

Text Microsoft Word - Anusuyah 4196-AAP.pdf - Published Version Restricted to Repository staff only Download (3MB)

Abstract

Customer retention is a key competitive advantage in the telecommunications industry, achieved by aligning service offerings with evolving customer expectations. This study aims to develop a predictive framework that integrates customer digital demography, Telco offerings, and behavioral change metrics to anticipate three customer decisions: to stay with current offerings, change offerings within the same provider, or churn to a competitor. The dataset comprises a time-series structure simulating 5 customers across 7 periods. Each record includes 7 demographic variables, 5 Telco offering attributes, and 5 change metrics, generating a multidimensional profile of evolving customer behavior. A Change Metric, computed on a 0–1 scale, was applied to quantify dissatisfaction. Thresholds were set to indicate likely outcomes: >0.50 for churn, >0.40 for change within provider, and ≤0.40 for retention. A Multi-Layer Perceptron model was employed to train and validate predictive outcomes. After the initial setup, the model was iteratively updated across time periods, with predictions from one cycle informing simulated service adjustments in the next. The model achieved a significant predictive rate (β=0.959, t=41.15, p<0.001). Trends in Change Metrics provided early warning for churn, enabling proactive interventions. This study demonstrates the feasibility of AI-driven predictive models for customer behavior analysis in dynamic environments. The framework is scalable, requires no complex infrastructure, and can be adapted across varied Telco settings. Future work may focus on validating the model with realworld datasets and expanding it to incorporate real-time service adjustment mechanisms.

Item Type:	Article
Uncontrolled Keywords:	Machine learning, artificial intelligence.
Subjects:	H Social Sciences > HD Industries. Land use. Labor > HD28-70 Management. Industrial Management > HD30.2 Electronic data processing. Information technology. Including artificial intelligence and knowledge management
Divisions:	Faculty of Management (FOM)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	05 Oct 2026 04:21
Last Modified:	05 Oct 2026 04:21
URII:	http://shdl.mmu.edu.my/id/eprint/16873

Downloads

Edit (login required)