Open Access

ARTICLE

Enhancing Intrusion Detection Systems Using Hybrid AI-Based Approaches

Mohammad Alshinwan¹, Radwan M. Batyha^1,2, Walaa Alayed^3,*, Saad Said Alqahtany⁴, Suhaila Abuowaida⁵, Hamza A. Mashagba⁶, Azlan B. Abd Aziz^6,*, Samir Salem Al-Bawri⁷

1 Faculty of Information Technology, Applied Science Private University, Amman, Jordan
2 Department of Computer Science, Faculty of Science and Information Technology, Irbid National University, Irbid, Jordan
3 Department of Information Technology, College of Computer and Information Sciences, Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia
4 College of Computer and Information Systems, Islamic University of Madinah, Madinah Munawarah, Medina, Saudi Arabia
5 Department of Data Science and Artificial Intelligence, Faculty of Information Technology, Al al-Bayt University, Mafraq, Jordan
6 Centre for Wireless Technology (CWT), Faculty of Engineering and Technology, Multimedia University, Melaka, Malaysia
7 Space Science Centre, Institute of Climate Change, Universiti Kebangsaan Malaysia (UKM), Bangi, Malaysia

* Corresponding Authors: Walaa Alayed. Email: ; Azlan B. Abd Aziz. Email:

(This article belongs to the Special Issue: Artificial Intelligence Methods and Techniques to Cybersecurity)

Computers, Materials & Continua 2026, 87(2), 97 https://doi.org/10.32604/cmc.2026.072806

Received 04 September 2025; Accepted 08 January 2026; Issue published 12 March 2026

Abstract

Safeguarding modern networks from cyber intrusions has become increasingly challenging as attackers continually refine their evasion tactics. Although numerous machine-learning-based intrusion detection systems (IDS) have been developed, their effectiveness is often constrained by high dimensionality and redundant features that degrade both accuracy and efficiency. This study introduces a hybrid feature-selection framework that integrates the exploration capability of Prairie Dog Optimization (PDO) with the exploitation behavior of Ant Colony Optimization (ACO). The proposed PDO–ACO algorithm identifies a concise yet discriminative subset of features from the NSL-KDD dataset and evaluates them using a Support Vector Machine (SVM) classifier. Experimental analyses reveal that the PDO–ACO model achieves superior detection accuracy of 98% while significantly lowering false alarms and computational overhead. Further validation on the CEC2017 benchmark suite confirms the robustness and adaptability of the hybrid model across diverse optimization landscapes, positioning PDO–ACO as an efficient and scalable approach for intelligent intrusion detection.

---

Keywords

---