Network Threat Monitoring System Using Machine Learning
Sai Yohathaa Sri R G, N S Rithika, M. Jamuna Rani · 2025
In the current digital era, where cyberattacks are increasing rapidly, safeguarding networks from advanced threats is essential. Continuous monitoring and detection are vital to prevent data breaches, system disruptions, and financial losses. This paper introduces a Network Threat Monitoring System (NTMS) which continuously monitors network traffic going through PCs, servers, and office networks, detecting suspicious activities and possible cyber-attacks in real-time. It informs about intrusion detection in good time, so that administering authorities can take quick preventive measures for securing vital infrastructure and data. The system has ensemble learning to enhance its detection accuracy. It involves integrating several algorithms under machine learning categories like Gradient boosting and random forest in a way that identifies both known and new threats. Its stacking ensemble applies a meta-model processing the output of individual classifiers to enhance its capability to detect threats. This multi-layer architecture, thus, overcomes the limitations of standalone models for higher accuracy with fewer false positives. The proposed system has achieved better performance considering the real-time traffic and network security analysis from the NSL-KDD dataset. Such developments are envisaged to create adaptive and robust solutions against modern cybersecurity challenges.