Research on application of Bayesian discriminant method in intrusion detection model

摘要

An intrusion detection model based on bayesian discriminant method is constructed in this paper. The intrusion detection problem is transformed into a discriminant classify problem in the model. We use stepwise discriminant method in selecting significant discriminatory variables to decrease computational complexity. The model is trained and validated by using the denial of service (DoS) attack data of KDD CUP99's data set. The result of the experiment shows high correctly identified percentage of the model. Intrusion detection system can be constructed with the model when there are sufficient network packet records as a support.