International Journal of Computer Techniques Volume 12 Issue 4 | Deep Neural Network Model for Customer Attrition Forecast in a Telecommunication Company
Deep Neural Network Model for Customer Attrition Forecast in a Telecommunication Company
Authors:
Emmah, Victor Thomas – victor.emmah@ust.edu.ng
Ordu, Princewill Okey – princewilordu@yahoo.co.uk
Bennett, Emmanuel O – okonni.bennett@ust.edu.ng
Department of Computer Science, Rivers State University, Nigeria
Journal: International Journal of Computer Techniques (IJCT)
Volume: 12 | Issue: 4 | Publication Date: July – August 2025
ISSN: 2394-2231 | Journal URL: https://ijctjournal.org/
Abstract
This study presents a Deep Neural Network (DNN) model for forecasting customer attrition in the telecom industry. Using historical data and a DevOps-driven ML pipeline, the model was trained with five layers and optimized using stochastic gradient descent. After 100 epochs, it achieved 98.1% accuracy, demonstrating strong predictive performance and offering actionable insights for customer retention strategies.
Keywords
Churn, DNN, Dataset, Epochs, Prediction, Optimization
Conclusion
The DNN model effectively predicts customer churn, supporting proactive retention strategies. Through iterative refinement, data preprocessing, and SGD optimization, the model delivers high accuracy and precision. The research provides a scalable framework for telecom providers to enhance customer satisfaction and reduce attrition.
References
Includes 20+ references from IEEE, IJACSA, Scientific Reports, Procedia Computer Science, and other peer-reviewed sources covering churn modeling, DNN architecture, and telecom analytics.
