一种基于主成分分析和模糊聚类的入侵检测方法  

A Method of Intrusion Detection Based on Principle Component Analysis and Fuzzy Clustering

作　　者：齐惠英[1] 

机构地区：[1]天津交通职业学院信息管理中心,天津300110

出　　处：《科技通报》2012年第12期51-53,共3页Bulletin of Science and Technology

基　　金：天津市教育科学"十二五"规划课题(VE4035)

摘　　要：入侵检测数据集中含有大量高维数据和冗余信息,降低了数据挖掘过程的系统学习效率与响应速度。提出一种基于主成分分析和模糊聚类相结合的入侵检测方法PCA-FC,通过对高维数据的预处理及特征提取,减少样本数据维数,然后执行数据聚集的方法提取出评价规则,从而有效地减少了训练数据的变量和干扰项,提高了分类效率。There exists lots of high dimensional data and redundancy information in intrusion detection data set,which will decrease the system study efficiency and response speed in data mining process.We propose an algorithm of intrusion detection PCA-FC based on the integration of orinciple component analysis and fuzzy clustering.It can reduce the sample data＇s dimensional by the pretreatment and feature extraction of high dimensional data.Then the evaluation rules is extracted by data aggregation.PCA-FC is testified to minimize the variables and disturbance terms of training data effectively and improve the classification efficiency.

关 键 词：入侵检测 主成分分析 聚类 特征抽取 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]