Intrusion detection using classification via clustering

Abstract: In today�s world there is widespread use of internet. It hence becomes a necessity for securing this access to the data that is stored on theworld wide web . Intrusion detection system is one such mechanism for detecting the intrusive patterns from the traffic patterns on the network . Datamining and statistical data analysis are some ways to detect these attacks. In this paper, we have presented a novel technique of intrusion detection where is classification is done on the results one gets after clustering the data set KDD '99 . The results obtained here are better than directly performing classification or clustering.

Keywords: Intrusion detection , data mining , statistical analysis , KDD '99