一种基于独立分量分析和Naive Bayesian网络的入侵检测方法

An New Intrusion Detection Method Based on ICA Model and Naive Bayesian Network

文章将独立分量分析穴ICA雪模型引入入侵检测系统,提出了基于独立分量分析和NaiveBayesian网络的入侵检测分类的新方法。通过把样本投影到有独立分量分析所确定的特征空间,来提高贝叶斯网络的分类性能,从而提高了入侵检测系统的性能。实验结果表明,这种基于独立分量分析模型的分类器具有良好的分类性能。

This paper applied the method of independent component analysis to intrusion detection system. An new intrusion detection method based on independent component analysis and Naive Bayesian Network is proposed. It can improve the cluster capacity by projecting the sample to the character space defined by independent component analysis. The experiment shows that the classifier based on independent component analysis model has good capacity.