Predictive Threat Evaluation in Complex IT Systems
COTS Security Incident and Event Management (SIEM) Systems process log events based on built-in rules and identify actionable incidents. These primarily identify known attacks. Using Machine Learning techniques such as Naive Bayes and AdaBoost algorithms, we aim to predict new attacks probabilistically for wired and wireless networks. The Machine Learning-based prediction system in tandem with an SIEM system to predict an attack before it actually occurs. Evaluate the effectiveness of the ML system comparing with the SIEM system in network attack prediction
