Abstrakt

Review on Data Mining Techniques for Intrusion Detection System

Sandeep D, M. S. Chaudhari

With the rapid growth of computer networks during the past few years, security has become a crucial issue for modern computer systems. A good way to detect illegitimate use is through monitoring unusual user activity. This can be achieved with an Intrusion Detection System, which identifies attacks and reacts by generating alerts or by blocking the unwanted data/traffic. These systems are mainly classified as Anomaly based Intrusion Detection Systems and Misuse based Intrusion Detection Systems. Anomaly based Intrusion Detection System has the benefit of detecting novel attacks but has a high false positive rate. On the other hand, Misuse based systems are signature based having higher accuracy. Misuse based Intrusion Detection System fails to detect novel attacks. To overcome these limitations, both Anomaly based and Misuse based Intrusion Detection Systems should be combined to form a new Hybrid Intrusion Detection System. A new Hybrid Intrusion Detection System is proposed. In this system, fuzzy data-mining concept based on genetic algorithm is used as an intrusion detection system. KDD dataset is used to train the system and test the system.