Enhancing rare attack detection in network intrusion systems through data augmentation

Key Points

• The central aim is to enhance the detection of rare attacks in network intrusion systems using data augmentation techniques.
• Utilized a machine learning framework for attack detection.
• Applied data augmentation strategies to improve model training.
• Analyzed model performance pre- and post-augmentation.
• Achieved a 25% increase in detection rates for rare attacks compared to baseline measurements.
• Demonstrated improved model accuracy with a 0.75 F1 score after data augmentation.
• Reduction in false positives by 15% when compared to previous detection methods.