摘要
提出一种基于数据挖掘和混合遗传算法(HGA)的自适应模型生成(AMG)模型。采用改进的聚类算法,从网络和系统的行为记录中划分出正常/异常行为库,利用HGA从行为库中挖掘出入侵规则加入规则库中,通过混合检测模块进行检测。实验结果证明,该AMG模型能以更高的检测率、更低的误检率检测未知的网络入侵。
The paper proposes an Adaptive Model Generation(AMG) model based on Hybrid Genetic Algorithm(HGA) and data mining.An improved clustering algorithm is used to divide the normal/abnormal behavior library that is from network and system behavior records.HGA is used to dig out the invasion rules,the rules are put into the rule base,and detection is realized by hybrid detection module.Experimental results show that the AMG model can detect unknown intrusion with higher detection rate and lower false detection rate.
出处
《计算机工程》
CAS
CSCD
2012年第7期99-101,共3页
Computer Engineering
基金
福建省仿脑智能系统重点实验室开放课题基金资助项目(BLISSOS2010103)
关键词
数据挖掘
入侵检测
混合遗传算法
自适应模型生成
聚类算法
信息增益
data mining
intrusion detection
Hybrid Genetic Algorithm(HGA)
Adaptive Model Generation(AMG)
clustering algorithm
information gain