摘要
高校网络开放性较强,导致入侵特征的多样性,很难建立统一的训练入侵检测模型,造成高校网络入侵检测存在弊端。本文提出一种基于无监督聚类关联的优化入侵检测算法,在建立的关联网络中对数据进行聚类分析,运用分层聚类方法分析网络,完成数据模型的建立。仿真实验表明,这种无监督入侵检测模型方法,克服了高校网络外网访问数据的识别特性不明显的问题,提高了高校网络入侵检测的准确率,取得了满意的结果。
University network openness is stronger,lead to the diversity of intrusion characteristics,it is difficult to establish unified training intrusion detection model,cause the network intrusion detection has shortcomings.This paper presents a based on unsupervised clustering associated optimization intrusion detection algorithm,in establishing connection network for data clustering analysis,using hierarchical clustering method analysis network,the establishment of the complete data model.Simulation experiments show that this kind of unsupervised intrusion detection model method,overcome the university network external network to access data identification characteristics is not obvious,improve the college network intrusion detection accuracy,and satisfactory results were obtained.
出处
《科技通报》
北大核心
2013年第6期95-97,共3页
Bulletin of Science and Technology
关键词
高校网络
入侵检测
无监督
the network
intrusion detection
without supervision
