基于熵的实值免疫检测器降维优化算法  

Real-valued Immune Detector Dimension-reducing Optimization Algorithm with Entropy

作　　者：孙永倩[1] 席亮[1] 

机构地区：[1]哈尔滨理工大学计算机科学与技术学院,黑龙江哈尔滨150080

出　　处：《哈尔滨理工大学学报》2014年第2期68-72,共5页Journal of Harbin University of Science and Technology

基　　金：国家自然科学基金(61103149)

摘　　要：针对免疫机制的入侵检测系统中实值检测器的维数灾难问题,借鉴信息熵理论,提出一种基于熵的实值检测器降维优化算法,使检测器在保留足够信息量的前提下在低维空间中运行,有效提高检测效率.实验表明,降维后的检测器表现出更好的检测性能.In order to deal with the ＂dimension curse＂ problem of real-valued detector in immunity-based intrusion detection system,we borrowed ideas from the information entropy theory,and a real-valued immune detector dimension-reducing optimization algorithm is proposed in this paper.The optimized detectors can retain the most useful information and work in the low-dimensional shape-space,which can improve the detection efficiency.The experiments show that the dimension-reduced detectors have better detector performances.

关 键 词：人工免疫 入侵检测 检测器 降维 熵 

分 类 号：TP393.0[自动化与计算机技术—计算机应用技术]