Hybrid Intrusion Detection System Using K-means and J-48

Manvir Singh, Jatinder Singh Bal
2017 International Journal of Security and Its Applications  
Security attacks through internet are increased continuously in recent years therefore the Information security is globally very serious problem at the present time. The importance of protecting networks is most important. The need of intrusion detection system has been brought out. In this work a hybrid intrusion system based on data mining techniques using Kmeans and J48 is proposed. In this paper work hybrid intrusion detection system is proposed first Misuse based detection model is build
more » ... ing K-means clustering algorithm then based upon the output of misuse detection phase Anomaly based intrusion detection model is build using improved J-48 classification algorithm in this way the combine output of these two phases attains very higher detection accuracy. For training and testing of hybrid intrusion detection model famous NSL-kdd dataset is used.
doi:10.14257/ijsia.2017.11.1.02 fatcat:nh2c5duplzbqfd6j6zhlo3eziu