现代电子技术2012，Vol.35Issue(19)：73-75,81,4.
基于主成分分析的改进贝叶斯网络入侵检测研究
Bayesian network intrusion detection technology based on principal component analysis and sliding window
摘要
Abstract
In the traditional Bayesian network intrusion detection technology, it is not taken into account that the intrusion detection data set has an excessive number of attributes, which leads to an excessive calculation in the process of Bayesian network structure, and greatly affect the detection efficiency. In addition, the traditional Bayesian network intrusion detection technology does not consider the current attacks and security status in the detection process, and fixed Bayesian network generated by the original training data set is used to test the new data set, which has a certain impact on detection accuracy. To solve the abovementioned two problems, a new Bayesian network intrusion detection technology based on principal component analysis and the sliding window is given. Principal component analysis method is adopted to reduce data dimensionality, and a sliding window mechanism and the detected test data are used to update the training data set, which can reflect the current security status of the system. The experiments show that improved technology can greatly reduce the data dimension, and improve the detection accuracy.关键词
特征选择/主成分分析/滑动窗口/贝叶斯网络/入侵检测Key words
feature selection/ principal component analysis/ sliding window/ Bayesian network/ intrusion detection分类
信息技术与安全科学引用本文复制引用
冯祖洪,李静..基于主成分分析的改进贝叶斯网络入侵检测研究[J].现代电子技术,2012,35(19):73-75,81,4.基金项目
宁夏高等学校科研项目(2011JY008) （2011JY008）
