A Unified Approach for Outlier Detection Using Change Point for Intrusion Detection
The number of computers connected to a network is increasing and Network Intrusion Detection System (NIDS) is one of the essential to an organization. The challenge here is to react quickly to different events with high detection rates and low false positives for real time traffic. Current research is in the problem of discovering change points at which properties of time-series data changes using change point detection algorithms. Outlier detection is the problem of discovering abnormal or deviating data points with respect to distribution in time-series data. Most previous works address detection of outliers and change points separately. In this paper, a novel approach for outlier detection and change points is used.