A Novel Network Traffic Anomaly Detection Approach Using the Optimal $\varphi$-DTW

摘要: Under the current severe situation of cyber security, it is great significance to propose an effective anomaly detection approach for ensuring stability network. It generally known that network traffic data a kind typical streaming time series data, which are recorded by equipments usually accompanied instants. In order detect anomalous sections in effectively, we unsupervised based on definition utilizing optimal $\varphi$ -DTW and corresponding similarity matrix, called ADOPD. Comprehensive experiments have demonstrated our proposed achieves satisfying performance detecting real world sets.