一个基于遗传分类器的入侵检测系统  

Intrusion detection system based on genetic classifier

作　　者：关健[1] 刘大昕 

机构地区：[1]哈尔滨工程大学计算机科学与技术学院,黑龙江哈尔滨150001

出　　处：《哈尔滨工程大学学报》2003年第2期201-204,共4页Journal of Harbin Engineering University

摘　　要：为了解决计算机系统中入侵检测问题,提出基于遗传分类器的方法,给出了编码方案和规则匹配算法,进而改进了规则更新算法.依据这些算法构建的系统经过仿真实验得到较好的结果,对个别攻击检测率可以达到96%.基于遗传分类器的入侵检测系统具有多级别特征采集和入侵规则集自动更新的特点,适合用于计算机安全防御领域.For detection of intrusions in computer systems,a genetic classifierbased intrusion system was proposed with encoding scheme and rulematching algorithm given,thereby improving the rule updating algorithm.Simulation tests show that the systems established using these algorithms can be used to achieve satisfactory results,and the detection probability can go up to 96% for some intrusions.The genetic classifierbased intrusion detection system proposed features multilayer information acquisition and automatic updating of intrusion rule sets and is therfore applicable in the computer defense domain.

关 键 词：入侵检测 遗传算法 分类器 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]