Network Congestion Tracking and Detection in Banking Industry Using Machine Learning Models

Abstract

The escalating threat of congestion in wireless networks on a global scale prompts the need for effective detection and management techniques. This study investigates the tracking and detection of congestion in wireless networks, particularly within the banking industry, where digital transactions are rapidly increasing. It addresses the challenge of congestion management through machine learning (ML) models, aiming to enhance network performance and service quality. This research evaluates various ML algorithms, including Support Vector Machines, Decision Trees, and Random Forests, to identify the most effective approach for congestion detection. This research utilizes a dataset sourced from MainOne Limited, which covered August 18th, 20th, 22nd, 23rd, and 24th, 2023, and included banking operation hours from 7 AM to 4 PM each day. Preprocessing of data is conducted to optimize model training. Following training, various performance metrics including accuracy, precision, recall, F1 score, response time, and confusion matrix are assessed. Results demonstrate that Random Forest outperforms other models in accuracy, precision, recall, F1 score, and response time, with an accuracy of 98.90%. This research discusses the importance of continuous innovation in banking network analytics to tackle evolving congestion challenges. Future recommendations include leveraging advanced ML techniques like deep learning and reinforcement learning and exploring ensemble learning methods to enhance congestion detection models further.