Muhammad Usama Nazir

A dynamic explainable AI framework for real-time intrusion detection and automated alert prioritization in security operation centers

The Dynamic Explainable Framework (DEF), which couples a machine-learning detection backbone with inference-time SHAP and LIME explanations and an explanation-aware alert prioritization layer, is proposed and designed and profiled for real-time operation on benchmark traffic traces.