摘要
基于模式识别方法的入侵检测系统首先要解决的一个问题就是特征选择,该文依据数据分布和相关分析两方面,提出了一种基于有监督学习的特征选择方法。根据实验结果可以看出,该算法执行效果较好,且时间复杂性较低。
The first problem that needs to be solved in intrusion detection system based on pattern recognition method is feature selection. This paper proposes a method for feature selection in supervised learning depending on data distribution and correlation analysis. According to the experimental results, this algorithm performs well and it's time complexity is low.
出处
《计算机工程》
EI
CAS
CSCD
北大核心
2005年第13期22-23,45,共3页
Computer Engineering
基金
教育部跨世纪人才基金资助重点科研项目(02029)
关键词
特征选择
相关分析
有监督学习
入侵检测
Feature selection
Correlation analysis
Supervised learning
Intrusion detection
