A DeepSHAP-Based Adversarial Attack on Machine Learning-Based Network Intrusion Detection

The primary objective of this project is to develop a robust Network Intrusion Detection System (NIDS) using machine learning techniques and evaluate its vulnerability to adversarial attacks through DeepSHAP analysis. The system utilizes Cascade Forest, Probabilistic Neural Network (PNN), and Echo State Network (ESN) models to accurately classify network traffic as normal or malicious. DeepSHAP is employed to identify the most influential features contributing to model predictions and to generate adversarial samples by manipulating these critical features. The project aims to analyze the impact of adversarial attacks on intrusion detection performance and compare the robustness of different machine learning models. Additionally, the system focuses on improving detection accuracy, reducing false alarms, and enhancing the security and reliability of machine learning-based intrusion detection systems