Adaptive Ensemble Classification with Recursive Feature Selection for Network Intrusion Detection

Abstract

The growth of cyber threats demands a robust and adaptive intrusion detection system (IDS) capable of effectively recognizing malicious activities from network traffic. However, the existing class imbalance in network data presents a significant challenge to traditional IDS. To address this issue, feature selection (FS) techniques are applied, such as recursive feature elimination (RFE) to reduce the dimensionality of the data and enhance the performance of the intrusion detection models. A novel approach is proposed that leverages an ensemble method, integrating the strengths of multiple classifiers to effectively manage class imbalance and improve detection accuracy. The model is evaluated on the CIC-IDS 2017 dataset, which includes a variety of network traffic scenarios and attack patterns. After applying advanced preprocessing and FS techniques, the ensemble model achieves high performance, with the voting classifier (RF + DT) reaching an impressive accuracy of 99.5%. The results demonstrate the effectiveness of the proposed method in handling imbalanced data while maintaining high detection performance, making it a promising solution for enhancing the reliability and efficiency of IDS in real-world cyber defense scenarios