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 classifierbased intrusion system was proposed with encoding scheme and rulematching 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 classifierbased intrusion detection system proposed features multilayer information acquisition and automatic updating of intrusion rule sets and is therfore applicable in the computer defense domain.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
2003年第2期201-204,共4页
Journal of Harbin Engineering University
关键词
入侵检测
遗传算法
分类器
intrusion detection
genetic algorithms
classifier
