| 基于半监督分类的BGP异常检测 |
| |
| 引用本文： | 李庆强,魏振钢,孙笑非,马丽晶.基于半监督分类的BGP异常检测[J].计算机应用,2008,28(Z2). |
| |
| 作者姓名： | 李庆强 魏振钢 孙笑非 马丽晶 |
| |
| 作者单位： | 中国海洋大学信息科学与工程学院,山东,青岛,26100 |
| |
| 摘 要： | 异常边界网关协议(BGP)事件会影响网络的稳定性和可靠性,而网络环境下未标记样本较有标记样本容易获得,对此提出了基于半监督分类的异常检测框架.主要研究了高斯混合模型和直推式支持向量机,使用Slammer蠕虫相关BGP数据进行了实验,并对算法性能作了比较.实验证明半监督分类算法在BGP异常检测中切实可行.
|
| 关 键 词： | 半监督分类 异常边界网关协议 异常检测 高斯混合模型 直推式支持向量机 |
Anomaly detection in BGP based on semi-supervised classification |
| |
| LI Qing-qiang,WEI Zhen-gang,SUN Xiao-fei,MA Li-jing.Anomaly detection in BGP based on semi-supervised classification[J].journal of Computer Applications,2008,28(Z2). |
| |
| Authors: | LI Qing-qiang WEI Zhen-gang SUN Xiao-fei MA Li-jing |
| |
| Affiliation: | LI Qing-qiang,WEI Zhen-gang,SUN Xiao-fei,MA Li-jing(College of Information Science , Engineering,Ocean University of China,Qingdao Sh,ong 266100,China) |
| |
| Abstract: | Abnormal Border Gateway Protocol(BGP) events can worsen the stabilization and reliability of networks.An anomaly detection framework based on semi-supervised classification was introduced which was suited for such a situation that gaining labeled data was more difficult than unlabeled data.Gaussian Mixture Models(GMM)and Transductive Support Vector Machines(TSVM) were studied,and the BGP data concerning the Slammer worms were used in the experiments,and the performances were compared.The results show that s... |
| |
| Keywords: | semi-supervised classification Border Gateway Protocol(BGP) anomaly detection Gaussian Mixture Models(GMM) Transductive Support Vector Machines(TSVM) |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
