Multi-class imbalanced learning implemented in network intrusion detection

In view of the problems of using traditional machine learning method to detect the network intrusions, a network intrusion detection model based on multi-class imbalanced learning is proposed. Based on the consideration that there is within-class imbalance in large data sets and multi-class data sets, every class of the training data is firstly clustered. Some minimum bounding hyperspheres are formed by Support Vector Date Description (SVDD) according to the clustering results. A test sample is assigned the label of hyperspheres if its distance to the sphere center is smaller than or equal to the radius. The model is testified by experiments on network security data sets.