Citation

Alsadawi, Saja U H and Foo, Yee Loo (2025) AI-empowered Network Intrusion Detection System. In: 2025 Multimedia University Engineering Conference, MECON 2025, 21 July 2025 - 23 July 2025, Cyberjaya, Malaysia.

Text IEEE Xplore Full-Text PDF_23.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

As cyber threats continue to evolve, traditional defence mechanisms like firewalls and signature-based solutions continue to lose efficacy. Machine learning-based NIDS present a promising alternative but also face issues of high-dimensional and imbalanced data and high false positive rates. This work addresses these issues with the use of preprocessing techniques like SMOTE, PCA, and XGBoost-based feature selection, and comparison of multiple machine learning models on the CSE-CIC-IDS2018 dataset. Experimental results show that K-Nearest Neighbors (KNN) model achieved the lowest false alarm rate (0.0053) in binary classification and Random Forest performed best in multiclass classification with 91.90% accuracy and 0.9292 F1-score. Compared to recent studies, our approach achieved higher precision and accuracy with simpler models and shorter training duration, showing its suitability for real-world NIDS deployment.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Intrusion detection, NIDS, machine learning,
Subjects:	Q Science > Q Science (General) > Q300-390 Cybernetics
Divisions:	Faculty of Artificial Intelligence & Engineering (FAIE)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	17 Mar 2026 06:42
Last Modified:	17 Mar 2026 06:42
URII:	http://shdl.mmu.edu.my/id/eprint/15494

Downloads

Edit (login required)