Privacy-aware intrusion detection system

Abstract

Intrusion Detection System (IDS) are vital for monitoring computer system and networks to detect unauthorized access or attacks. As cyber threats from insiders and outsiders become more sophisticated, safeguarding digital information is crucial. IDS use two main methods: Signature-based (SIDS), which detect known attack patterns, and Anomaly-based (AIDS), which identify unusual activities. By analyzing network traffic for suspicious behavior, IDS quickly raise alerts and log detailed records for investigation, maintaining network integrity and confidentiality. In this study, we evaluated the effectiveness of various machine learning models for IDS. We tested Support Vector Machine (SVM), k-Nearest Neighbor (kNN), Random Forest, and eXtreme Gradient Boosting (XGBoost) with a dataset of attack types. Random Forest outperformed the other models, achieving an accuracy of 99.9958%. Its ability to handle many features and avoid overfitting makes it highly suitable for IDS.