An Effective Parallel SVM Intrusion Detection Model for Imbalanced Training Datasets

摘要

In the field of network security, the Intrusion Detection Systems (IDSs) always require more research on detection models and algorithms to improve system performance. Meanwhile, higher quality data is critical to the accuracy of detection models. In this paper, an effective parallel SVM intrusion detection model with feature reduction for imbalanced datasets is proposed. The model includes 3 parts: 1) NKSMOTE-a Modified unbalanced data processing method. 2) feature reduction based on Correlation Analysis. 3) Parallel SVM algorithm combining clustering and classification. The NSL-KDD dataset is used to evaluate the proposed method, and the empirical results show that it achieves a better and more robust performance than existing methods in terms of the accuracy, detection rate, false alarm rate and training speed.