A prediction model for network traffic anomaly detection

Abstract

In order to improve the defects in intrusion detection system (IDS) on network traffic anomaly detection (NTAD). A new IP based NTAD was proposed, which connected with linear regression analysis and IP adjust. This mechanism point at the features of the network, use linear regression analysis on the IP data, this can predict the abnormal traffic more accurately. The experiment results also showed the new mechanism have more accuracy and higher recall ratio.