Network Intrusion Detection and Anomaly Classification

A software application known as an intrusion detection system (IDS) is utilized to monitor networks for any unauthorized or malicious activities that may violate security policies related to system confidentiality, integrity, and availability. Our thesis involved conducting extensive literature reviews on various types of IDS, anomaly detection methods, and machine learning algorithms that can be utilized for detection and classification. IDS are crucial for safeguarding advanced communication networks and were initially designed to identify specific patterns, signatures, and rule violations. In recent years, machine learning and deep learning approaches have been employed in the field of network intrusion detection, providing promising alternatives. Our proposed system involved identifying noisy features through causal intervention, preserving only those features that had a causality with cyberattacks. The ML algorithm was then utilized to make a preliminary classification to select the most relevant types of cyberattacks, ultimately leading to the detection of unique labeled cyberattacks through the counterfactual detection algorithm.