Classifier Selection Model for Network Intrusion Detection Using Data Mining

Intrusion detection is one of the core technologies of computer security. It is required to protect the security of computer network systems. Due to large volumes of security audit data as well as complex and dynamic properties of intrusion behaviors, optimizing performance of network intrusion detection becomes an important open problem. Data mining techniques are being applied in building intrusion detection systems to protect computing resources against unauthorized access. Classification methods are employed to categorize the network attacks. Usually one classifier algorithm is used to classify various network attacks. In this paper, we evaluated the performance of a comprehensive set of classifier algorithms using KDD99 dataset. It is found that certain algorithms perform better for certain attack classes. So a set of classifier algorithms namely, BayesNet, Naive Bayes, J48, Decision Table, JRip, OneR, IBk are compared based on the 10-fold cross validation test. Finally, two classifier algorithm selection models are proposed based on the evaluation results to categorize the network attacks.