Hybrid Ant colony-Cuttlefish optimization for feature selection in machine learning-based intrusion detection systems
Intrusion Detection Systems (IDS) constitute a critical component of contemporary cybersecurity frameworks; however, their performance is frequently hindered by the high dimensionality of network traffic data, which contributes to increased false alarm rates and computational inefficiencies in conventional feature selection approaches. To address these limitations, this study proposes a novel hybrid feature selection framework that integrates the Ant Colony Optimization (ACO) and the Cuttlefish Algorithm (CFA). The hybrid approach leverages the global search exploration capability of ACO alongside the local search refinement strength of CFA to improve feature optimization for intrusion detection tasks. The proposed ACO-CFA algorithm was extensively evaluated using the KDD Cup 99 benchmark dataset, while the effectiveness of the optimized feature subsets was assessed through three machine learning classifiers: Random Forest (RF), Decision Tree (DT), and Support Vector Machine (SVM). Experimental findings revealed outstanding classification performance across multiple data partitioning ratios. Among the evaluated models, the Random Forest classifier consistently demonstrated superior effectiveness, achieving an accuracy of up to 99.97% while maintaining an optimal balance between precision and recall. Although the DT classifier produced comparable accuracy with faster computational performance, the SVM exhibited substantially higher computational costs despite its strong predictive capability. The study concludes that the hybrid ACO-CFA framework provides an efficient and scalable solution for feature selection, significantly improving IDS detection accuracy while minimizing computational complexity. Consequently, the proposed approach offers a robust foundation for the development of adaptive and high-performance intrusion detection systems capable of addressing increasingly sophisticated cyber threats.