ISSN: 2182-2069 (printed) / ISSN: 2182-2077 (online)
Enhancing Intrusion Detection Systems with XGBoost Feature Selection and Deep Learning Approaches.
In response to the growing complexity and frequency of cyber-attacks, this study investigates the efficacy of the XGBoost algorithm combined with deep learning (DL) techniques, such as ANN, 1DCNN, and BiLSTM, for creating accurate intrusion detection systems (IDSs). The synthetic minority oversampling technique (SMOTE) is employed to address data imbalance, and the proposed approach is evaluated against three IDS datasets: NSL-KDD, CIC-IDS2017, and UNSW-NB15. The high accuracy and low error rate of the classification models demonstrate the potential of the proposed approach in IDS design, with the ANN + BiLSTM model combined with SMOTE consistently outperforming other models in terms of precision and recall for individual attack classes. This emphasizes the importance of data balancing techniques and the effectiveness of integrating XGBoost and DL approaches for accurate IDSs. Future research can focus on implementing novel sampling techniques explicitly designed for IDSs to enhance minority class representation in public datasets during training.