Q-Detect: A Lightweight Quantum-Inspired Feature Selection Framework for Computationally Efficient Detection in High-Dimensional Networked Systems
DOI:
https://doi.org/10.29284/0tapqn58Keywords:
cyber security, Quantum, machine learning .Abstract
The current technical systems that run under resource limited environments demand smart computationally structures capable of providing high performance at minimal processing overhead. Wrapper-based feature selection algorithms like Particle Swarm Optimization (PSO), are used in high-dimensional data driven systems to increase accuracy in classifications, but due to its high reliance on repeated assessment of classifiers, it is highly computationally expensive. In this paper, the authors present a lightweight quantum-inspired feature selection model named Q-Detect that is based on Quantum-behaved Particle Swarm Optimization (QPSO). The suggested framework uses a simplified quantum position-update tool as well as a rigid evaluation budget, making it possible to search globally efficiently and minimizing computational load greatly. Experimental analysis over NSL-KDD benchmark shows that Q-Detect can make a maximum classification accuracy of 77.99 which is better than classical baselines such as Mutual Information, conventional PSO, and non-optimized models. During runtime analysis, it has been found that, whereas classical PSO takes 6003.54 seconds to perform 600 wrapper evaluations, Q-Detect lowers the overall optimization cost by 33-40 percent to 4002.36 seconds by limiting the evaluation procedure to 400 iterations. Further experiments using the large-scale CIC-IDS2017 data substantiate that the proposed framework is scalable and applicable to the large-scale, high-dimensional technical systems in the modern world. All in all, the findings have shown that Q-Detect is an effective and deployable quantum-inspired optimization framework that can be applied in computational intensive technical and intelligent systems..
References
[1] M. Ring et al., “A survey of network-based intrusion detection data sets,” Computers & Security, 2019.
[2] M. Ring, S. Wunderlich, D. Grüdl, D. Landes, and A. Hotho, “A survey of network-based intrusion detection data sets,” Computers & Security, vol. 86, pp. 147–167, 2019.
[3] S. Mahdavifar and A. Ghorbani, “Deep learning for intrusion detection: A survey,” IEEE Communications Surveys & Tutorials, vol. 21, no. 4, pp. 2821–2847, 2019.
[4] M. A. Ferrag, L. Maglaras, and H. Janicke, “Deep learning for cybersecurity: A survey,” IEEE Access, vol. 7, pp. 41585–41600, 2019.
[5] [4] R. Panigrahi and S. Borah, “A survey on feature selection methods,” Computer Science Review, vol. 40, p. 100411, 2021.
[6] R. Moustafa and J. Slay, “UNSW-NB15: A comprehensive dataset for network intrusion detection systems,” in Proc. IEEE MILCOM, 2015, pp. 1–6.
[7] S. Revathi and A. Malathi, “A comparative study of feature selection techniques in intrusion detection,” IEEE Access, vol. 9, pp. 110477–110487, 2021.
[8] N. Almusallam and M. Zolkipli, “Feature selection methods for intrusion detection systems: A review,” Algorithms, vol. 13, no. 12, p. 332, 2020.
[9] P. Sharma and R. Kumar, “A comprehensive review of feature selection techniques in machine learning for high-dimensional data,” Expert Systems with Applications, vol. 222, p. 119794, 2023.
[10] S. S. Sodhro et al., “Swarm-intelligence-based feature optimization for IoT security,” IEEE Access, vol. 7, pp. 127554–127563, 2019.
[11] X. Chen and S. Kumar, “Improved binary particle swarm optimization for high- dimensional feature selection,” Information Sciences, vol. 613, pp. 419–439, 2022.
[12] K. Yang, B. Li, Q. Wu, and X. Chen, “Genetic algorithm-based optimization for intrusion detection,” Computers & Security, vol. 102, p. 102162, 2021.
[13] Y. Zhang, H. Chen, and C. Yang, “Quantum-inspired optimization algorithms: A systematic review and applications,” Engineering Applications of Artificial Intelligence, vol. 114, p. 105089, 2022.
[14] F. Yan et al., “Quantum-inspired evolutionary algorithms: A review and applications,” Engineering Applications of Artificial Intelligence, vol. 109, p. 104657, 2022.
[15] J. Sun, H. Feng, and W. Xu, “Quantum-behaved particle swarm optimization: A survey and applications,” Expert Systems with Applications, vol. 213, p. 119067, 2023.
[16] Y. Huang and J. Zhao, “Quantum-behaved PSO for intrusion detection,” IEEE Transactions on Emerging Topics in Computing, vol. 9, no. 3, pp. 1220–1232, 2021.
[17] M. Li et al., “Quantum feature selection via QUBO formulation,” in Proc. IEEE MASS, 2022, pp. 419–427.
[18] H. Qi et al., “Quantum-inspired optimization with extreme learning machines for intrusion detection,” Journal of Supercomputing, vol. 80, pp. 14373–14392, 2024.
[19] A. Al-Hashmi, M. Alazab, and K. Shuaib, “Deep neural networks for modern intrusion detection,” Applied Soft Computing, vol. 124, p. 109075, 2022.
[20] K. Deb and A. Pratap, “Swarm and evolutionary algorithms for high-dimensional search problems,” Applied Soft Computing, vol. 96, p. 106681, 2020. Int. J.Adv.Sig.Img.Sci, Vol. 12, No. 2s, 2026 862
[21] A. Buczak and E. Guven, “A survey of data mining and machine learning methods for cyber security intrusion detection,” IEEE Communications Surveys & Tutorials, vol. 18, no. 2, pp. 1153–1176, 2016.
[22] G. Aceto et al., “Survey on machine learning and deep learning for network traffic analysis,” IEEE Communications Surveys & Tutorials, vol. 21, no. 3, pp. 2557–2594, 2020.
[23] P. Narang, S. R. Chowdhury, and A. Chattopadhyay, “Quantum machine learning for cybersecurity: A survey,” IEEE Access, vol. 11, pp. 23137–23163, 2023.
[24] M. Zaman and R. Karray, “Redundancy and feature correlation in IDS datasets,” Information Sciences, vol. 584, pp. 363–380, 2022.
[25] H. Kim and D. Park, “Evaluating filter-based feature selection in large-scale IDS,” Expert Systems with Applications, vol. 183, p. 115503, 2021.
[26] L. Sun et al., “An empirical study on redundant features in intrusion detection datasets,” Computers & Security, vol. 120, p. 102806, 2022.
[27] B. Alshamrani, “Performance analysis of PSO-based IDS feature selection,” Future Generation Computer Systems, vol. 125, pp. 345–356, 2021.
[28] T. Gupta et al., “Genetic algorithm optimization for cyber-attack classification,” Applied Intelligence, vol. 52, pp. 14823–14839, 2022.
[29] R. Pereira et al., “Limitations of swarm intelligence for high-dimensional security data,” Information Sciences, vol. 627, pp. 89–107, 2023.
[30] C. Wang, Y. Liu, and Z. Huang, “Behavior analysis of particle swarm optimization in high- dimensional search spaces,” Applied Soft Computing, vol. 116, p. 108362, 2022.
[31] M. Sotoodeh et al., “Convergence challenges in swarm optimization for cybersecurity,”IEEE Access, vol. 11, pp. 14562–14578, 2023.
[32] I. Sharafaldin, A. Habibi Lashkari, and A. Ghorbani, “Toward generating realistic intrusion detection datasets: CIC-IDS2017,” CIC Technical Report, 2017.
[33] S. Suhail et al., “A comprehensive review of CICIDS2017 for modern IDS research,” Computers & Security, vol. 130, p. 103708, 2024.
[34] D. Han and Y. Wang, “Quantum-inspired hybrid optimization algorithms,” Knowledge- Based Systems, vol. 260, p. 110152, 2023.
[35] Z. Liu et al., “Quantum behaviors in optimization: A review,” Swarm and Evolutionary Computation, vol. 82, p. 101284, 2024.
[36] P. Mukherjee, R. Singh, and S. Alam, “Applications of QPSO in big data analytics,” Expert Systems with Applications, vol. 219, p. 119602, 2023.
[37] S. Zolanvari, R. Jain, and N. Tabrizian, “Limitations of classical IDS datasets in modern networks,” IEEE Internet of Things Journal, vol. 6, no. 1, pp. 103–115, 2019.
[38] W. Zong et al., “Network flow-based attack detection using CIC-IDS2017,” Applied Soft Computing, vol. 136, p. 110129, 2023.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
