基于一种改进K-means算法的入侵检测系统研究  

Research on the Intrusion Detection System based on an Improved K-means Algorithm

作　　者：谢霖铨[1] 张思洁[2] 

机构地区：[1]江西理工大学应用科学学院,江西赣州341000341000 [2]江西理工大学理学院,江西赣州341000

出　　处：《河南科技》2015年第8期23-25,共3页Henan Science and Technology

基　　金：江西省自然科学基金(20151BAB217010)

摘　　要：为了提高入侵检测系统的实时性和可靠性,基于传统的K-means算法应用于入侵检测系统的有效性和由于随机初始聚类中心而存在的缺陷,基于此缺陷而改进为选取已经初始化的聚类中心距离乘积最大的点为初始聚类中心,并将其运用到入侵检测领域。这种方法可以有效检测出不同类型的入侵行为,并且在KDDCUP99数据集的实验中取得了预期的效果。In order to improve the real-time performance and reliability of the intrusion detection system, based onthe validity of traditional K-means algorithm applied to intrusion detection system and due to the defects of the ran-dom initial clustering center. Based on the defect and make improvement to select the highest point of the initial clus-tering center distance product as the initial clustering center, and apply it to the intrusion detection field. This meth-od can effectively detect different types of intrusion behaviors, and it has achieved the expected results in the KDDCUP99 data set.

关 键 词：入侵检测 数据挖掘 无监督聚类 最大距离积法 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术] TP311.13[自动化与计算机技术—计算机科学与技术]