AN ITERATIVE GROUP BASED ANOMALY DETECTION METHOD FOR SECURE DATA COMMUNICATION IN NETWORKS
Unknown authors
Abstract
The most important task in intrusion detection is to identify the exact location of outliers which results in low false positive rate in identification of different kinds of attacks. Based on the arrangements made in outlier detection methods, the proposed methods have to identify the type of attack or outlier on the other hand has to predict potential attacks which causes reduced system performance. In the event when the log information does not contain the attack related information, no caution is raised. Thus, the false positive rate can be reduced extremely near to zero detecting outliers accurately. After identification of intrusions in the network, a Group Based Anomaly Detection Method (GADM) is proposed to incorporate recognition procedures just as levels of intrusion detection framework which maps the cross over features of different nodes for detecting intrusions quickly and accurately. The proposed method performance levels are high when compared to the traditional methods.