一种改进聚类算法在入侵检测中的应用  

The Application of An Improved Clustering Algorithm in Intrusion Detection

作　　者：赵云[1] 顾健[1] 张笑笑[1] 

机构地区：[1]公安部第三研究所,上海200031

出　　处：《信息安全与技术》2012年第12期15-19,共5页

基　　金：国家科技重大专项(2012ZX03002011)资助

摘　　要：本文首先介绍入侵检测系统的基本结构和研究情况,然后介绍了K-means聚类算法的目标函数、算法流程;在总结K-means聚类算法存在的问题的基础上,提出了一种改进的聚类算法。该算法为基于数据挖掘的入侵检测的设计提供了相关可操作的理论依据。最后,通过模拟实验,证明了改进算法的有效性。In this paper, the basic structure and research situation of intrusion detection system are introduced firstly, and then the objective function and procedure of the K-means clustering algorithm are presented. Based on the summarized problems of K-means clustering algorithm, an improved clustering algorithm is proposed. The algorithm provides a related operable theoretical basis for intrusion detection design based on data mining. The simulated experiments prove the validity of the improved algorithm.

关 键 词：数据挖掘 入侵检测 聚类算法 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]