A novel hybrid hunger games algorithm for intrusion detection systems based on nonlinear regression modeling.

Area of Science:

• Cybersecurity
• Artificial Intelligence
• Network Intrusion Detection

Background:

• Increasingly complex cyber threats necessitate advanced Anomaly-based Intrusion Detection Systems (AIDSs).
• Existing AIDSs face challenges including high false alarm rates, outdated datasets, and insufficient accuracy.
• Artificial intelligence offers potential solutions for enhancing AIDS capabilities.

Purpose of the Study:

• To propose a novel intrusion detection system for efficient and accurate detection of diverse cyber-attacks.
• To address limitations of current AIDS methods, such as imbalanced data and feature selection.

Main Methods:

• Utilized Smote-Tomek link for data balancing and preprocessing of the CICIDS dataset.
• Developed a hybrid meta-heuristic algorithm combining Gray Wolf, Hunger Games Search (HGS), and genetic algorithm operators for feature selection and attack detection.
• Modeled network behavior using nonlinear quadratic regression optimized with the hybrid HGS algorithm.

Main Results:

• The proposed feature selection technique removed over 80% of irrelevant features.
• The hybrid HGS algorithm demonstrated superior performance compared to baseline algorithms.
• Achieved an average test accuracy rate of 99.17%, outperforming the baseline algorithm's 94.61%.

Conclusions:

• The novel hybrid AI-driven intrusion detection system effectively mitigates cyber threats with high accuracy.
• The proposed method offers a significant improvement over existing techniques for network security.
• The system successfully addresses challenges in data preprocessing, feature selection, and detection accuracy.