Methodology for Research and Determination of Complex Cyber Attacks in Corporate Networks

Cybersecurity professionals are valuable assets nowadays, considering the constantly emerging obscure security threats. However, the volume of security incidents in the modern corporate networks often surpasses the capacity of the teams managing the security stack, leading to incident fatigue within organizations. In this paper, we propose a methodology that leverages context-aware predictions based on historical incident data to ease the decision making process for the security teams. Using Learning Vector Quantization neural networks, our approach helps to efficiently differentiate between false positive and true positive incidents, providing with actionable insights, and reducing the time spent on irrelevant alerts.