基于线性表示的协同入侵检测模型  

Collaborative Intrusion Detection Mode Based on Linear Representation

作　　者：梅松青[1] 

机构地区：[1]广州医科大学信息管理与信息系统系,广州510182

出　　处：《现代计算机》2013年第22期12-15,共4页Modern Computer

摘　　要：针对现有入侵检测算法误报率较高和鲁棒性较差的问题,提出一种基于线性表示的协同入侵检测模型,按照网络协议产生3类基于K-均值聚类算法的检测代理,以协同的方式对其相应网络数据做出决策。实验对比结果表明,该协同模型具有较高的检测率和较低的误警率。For the problem of slow speed of traditional intrusion detection,presents collaborative intrusion detection mode based on linear representation. According to the network protocol, generates three detection agents based on K-means clustering algorithm, then makes decisions in a synergistic manner, the results of experimental comparison show that the collaborative model has a higher detection rate and low false alarm rate.

关 键 词：入侵检测 线性表示 协同模型 检测率 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]