An Improved Approach for Signature and Anomaly based Intrusion Detection and Prevention
A Shivarkar Sandip, Akshat Ajit, J Dange Bapusaheb · Computational intelligence · 2012
Intrusion detection systems (IDS) are developing very rapid in recent years. But most traditional IDS can only detect either misuse or anomaly attacks. In this paper, we propose a system that combining both misuse and anomaly attacks. Hybrid intrusion detection is a novel kind of model combining the advantages of anomaly detection and misuse detection. We design a new hybrid intrusion system based on conditional random fields. Experimental results for Signature based detection based on the KDD 1999 Cup dataset shows that the proposed model is promising in