Published Online:April 2026
Product Name:The IUP Journal of Computer Sciences
Product Type:Article
Product Code:IJCS040426
DOI:10.71329/IUPJCS/2026.20.2.44-52
Author Name:Yusuf Mohammed Idowu, Francisca Nnoyelum Ogwueleka and Fatimah Binta Abdullahi
Availability:YES
Subject/Domain:Engineering
Download Format:PDF
Pages:44-52
The paper investigates the causes and implications of network downtime in Nigerian banks using machine learning (ML) techniques. The study adopts a qualitative research approach, using network performance indicators such as latency, packet loss, bandwidth utilization, server CPU load, third-party service availability and system upgrade activities. Supervised ML models, including Logistic Regression, Decision Tree, Support Vector Machine (SVM), and Random Forest, were developed and evaluated using standard performance metrics such as accuracy, precision, recall, F1 score and ROC-AUC. The results reveal that network downtime in Nigerian banks is primarily driven by technical and third-party factors, with third-party service availability, network latency, packet loss and bandwidth congestion emerging as the most significant contributors. Among the evaluated models, the Random Forest classifier demonstrated superior performance, achieving the highest prediction accuracy and ROC-AUC score, thereby confirming its effectiveness in capturing complex, nonlinear relationships inherent in banking network environment. The study further highlights the operational and financial implications of downtime, including service interruptions, transaction failures, and reduced customer trust. The findings underscore the potential of ML-based predictive models as proactive tools for minimizing network downtime and enhancing system resilience within the Nigerian banking sector.
Technology has infiltrated every aspect of human life, achieving several complex tasks quickly.