Review on Clustering and Classification techniques in Intrusion Detection Systems

Abstract

In the modern cyber world, the proportion of security threats is cumulating every day, and many researchers and security specialists are focusing on IDS (Intrusion Detection Systems) and the patterns in recognizing the alerts / events to detect and prevent them. The researchers and security specialists believe that the IDS is best way to protect the network and information assets. Hence this paper is focused on various threats and potential of IDS with its patterns in detecting the alerts / events. Basically, IDS has three different styles in detection the threats: Signature-based Detection (SD), Anomaly-based Detection (AD), and Stateful Protocol Analysis (SPA). The main area where the researchers and security specialists focusing is on techniques and algorithms used for clustering and classification. This paper mainly supports in understanding and analysing the various patterns on clustering and classifying the previous alerts / events which mainly supports in detection of threats with accuracy. This review will help to increase the detection accuracy of the IDS by enhancing the clustering and classification techniques which supports efficient execution of IDS over the network