Research on Multi-granularity Intrusion Detection Algorithm Based onSequential Three-Way Decision
Yongli Geng, Yongzhong Li, Shipeng Zhang · Proceedings of the 2021 5th International Conference on Electronic Information Technology and Computer Engineering · 2021
Intrusion detection is one of the significant research directions in the field of network information security, which has received widespread attention in academia and industry. How to distinguish the nature of network behavior is the important content of intrusion detection research. In view of this, the paper proposes Multi-Granularity intrusion detection algorithm based on Sequential Three-Way Decision (S3WD-MG). Firstly, the feature sets of different granularity are obtained through auto-encoding; Secondly, the relevant identification information in the features extracted by the auto-encoder network will increase with training time, thus forming a multi-granularity feature set; Finally, based on the sequential three-way decision theory and the decision threshold, the most appropriate decision is made for network behavior. The experimental consequence on the NSL-KDD test set shows that, S3WD-MG has higher detection rate and stronger robustness than other model in intrusion detection.