Network intrusion detection model using wrapper based feature selection and multi head attention transformers.

Area of Science:

• Computer Science
• Cybersecurity
• Network Security

Background:

• Modern networked systems in healthcare, industry, and homes face increasing cyberattack risks.
• Traditional security systems struggle with the volume, diversity of devices, and evolving attack vectors.
• Existing intrusion detection methods, including machine learning, still face accuracy challenges.

Purpose of the Study:

• To develop a novel intrusion detection model with enhanced accuracy.
• To address the limitations of current security systems in identifying sophisticated cyber threats.
• To leverage advanced machine learning and deep learning techniques for robust network defense.

Main Methods:

• Utilized the UNSW-NB15 dataset for model training and evaluation.
• Implemented a wrapper-based feature selection technique with machine learning algorithms.
• Employed a Multi-Head Attention-based transformer for intrusion prediction on selected features.

Main Results:

• The proposed model demonstrated improved accuracy in intrusion detection.
• Feature selection effectively reduced the feature space while retaining critical information.
• Performance metrics including Accuracy, Precision, Recall, and F-1 score were utilized for evaluation.

Conclusions:

• The developed model offers a more accurate and efficient approach to network intrusion detection.
• Feature selection is crucial for improving the performance of deep learning models in cybersecurity.
• The model shows significant potential for enhancing the security of interconnected systems.