Hybrid Feature Selection for Anomaly Detection in IoT Network Intrusion Detection Systems
Advances in Science, Technology and Engineering Systems Journal · 2026-04 · 11 (2) · 17-29
DOI: 10.25046/aj110203 ↗Kurz gefasst
Fachartikel in der Zeitschrift Advances in Science, Technology and Engineering Systems Journal, 2026-04. Verfasst von Mya Soe Soe Moe und 1 weiteren Autor:innen. Die Arbeit nennt 30 Literaturangaben.
Abstract (Originalsprache)
The rapid growth of Internet of things (IoT) devices have heightened the need for effective Intrusion Detection System (IDS) to combat evolving cyber threats. The IoT networks has the security challenges due to the heterogeneous and high-dimensional nature of network traffic data, redundant features, and class imbalance which hinder detection accuracy and efficiency. Effective IDS requires robust feature selection mechanisms to enhance detection accuracy and reduce computational complexity. The research proposes a hybrid feature section method that combines filter and embedded techniques through a weighted scheme. Chi-square and Mutual Information scores are fused with a weighting mechanism and interests with Random Forest feature importance for anomaly detection in IoT environments. The proposed hybrid feature selection approach is evaluated on three benchmark IoT intrusion detection datasets, ACI-IoT2023, BoT-IoT, and TON-IoT datasets using five supervised classifiers: Random Forest, Decision Tree, K-Nearest Neighbour, Support Vector Machine, and Naïve Bayes classifiers. The study evaluated the proposed system for both binary and multiclass classification scenario. Experimental results demonstrated that the proposed feature selection with Random Forest classifier achieves the highest performance 99.93%, 99.98%, 99.01% in accuracy on each dataset respectively.
Zitieren
APA
Moe, M. S. S., & Oo, W. M. (2026). Hybrid Feature Selection for Anomaly Detection in IoT Network Intrusion Detection Systems. Advances in Science, Technology and Engineering Systems Journal, 11(2), 17-29. https://doi.org/10.25046/aj110203
Vancouver
Moe MSS, Oo WM. Hybrid Feature Selection for Anomaly Detection in IoT Network Intrusion Detection Systems. Advances in Science, Technology and Engineering Systems Journal. 2026;11(2):17-29. doi:10.25046/aj110203
BibTeX
@article{Moe2026,
author = {Moe, Mya Soe Soe and Oo, Win Mar},
title = {Hybrid Feature Selection for Anomaly Detection in IoT Network Intrusion Detection Systems},
journal = {Advances in Science, Technology and Engineering Systems Journal},
year = {2026},
volume = {11},
number = {2},
pages = {17--29},
doi = {10.25046/aj110203}
}
EUVIMED