SMOTE-NCL: A re-sampling method with filter for network intrusion detection
Yong Sun, Feng Liu · 2016
Network intrusion detection research using KDDCUP 99 dataset often encounters challenges that classifiers could not handle the problem of uneven distribution of attack categories. This is known as imbalanced classification. The Synthetic Minority Over-sampling Technique (SMOTE) can balance the data distribution, but the noisy and boundary data generated by over-sampling can also reduce the classification performance. In this paper, the SMOTE-NCL method is proposed, which can overcome the influence of noisy and boundary data. This paper compares the results of the proposed methods with the latest methods of intrusion detection. Experimental results also show that the new method has better performance than the existing SMOTE method in intrusion detection. It can improve the performance of intrusion detection, especially for the detection of minority attacks.