Cross-year cyber-attack detection and temporal generalisation: an explainable machine learning approach

Key Points

• This research aims to enhance the detection of cyber-attacks using machine learning techniques that generalize across different time periods.
• Utilized explainable machine learning algorithms to analyze cyber-attack patterns
• Examined data from multiple years to improve detection accuracy
• Focused on temporal generalization methods for broader applicability
• Achieved significant improvement in detection rates over traditional methods
• Demonstrated strong predictive performance across different time frames
• Provided insights into temporal patterns of cyber-attacks, enhancing response strategies