PERFORMANCE EVALUATION OF ADVANCED INTRUSION DETECTION SYSTEM

_______________________________________________________________________________________ Abstract An intrusion detection system (IDS) plays an important role in securing our computer networks. There are many techniques designed to help in detecting and preventing such attacks. Intrusion detection system (IDS) is used to observe unwanted action on network systems and individual computers. However due to the opposite effect of using IDS, using individual methods of IDS only misuse or anomaly attacks can be detected. This paper proposed a model that integrates both approaches signature and anomaly based on IDS to reduce obtained alerts and detects new attacks. False alert rate, area under ROC curve, accuracy, recall, F-score, and P-test are the parameters used to evaluate the effectiveness of hybrid IDS. This paper also considers KDD Cup 1999 information set keeping in mind the end goal to show the profits of the proposed intrusion detection system described in the KDD Cup record previously.