摘要
针对基于计算机免疫的入侵检测系统中所面临着"不完全Self集"的问题,设计了基于决策树的主从结构的Self集构造算法。将决策树引入到传统的否定选择算法中,通过决策树把经过免疫耐受淘汰后的候选检测器进行重新分类,并将满足设定条件的候选检测器集合构造"从Self集",实现Self集的动态扩充,最后利用"匹配矛盾"淘汰"从Self集"中不合格的元素。实验分析结果表明了该算法的有效性,改善了检测器识别性能。
Aimed at solving the problem of＂self-set incomplete＂ that exists in intrusion detection system based on computer immune, a construction algorithm ofprincipal and subordinate structure self-set based on decision trees is designed. The decision trees are introduced to traditional negative-selection algorithm and the candidate detectors which have been eliminated by the immune tolerance are reclassified by the decision trees, and the candidate detectors that meet the setting conditions compose the ＂subordinate self-set＂ so as to achieve the dynamic expansion of the self-set. The unqualified elements in ＂subordinate self-set＂ are eliminated according to the ＂match conflict＂ method. Experimental results show that this algorithm is effective and improve the recognition performance of the detectors.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第7期1616-1618,1643,共4页
Computer Engineering and Design
基金
湖北省教育厅重点科研基金项目(2004D006)
关键词
计算机免疫
入侵检测
否定选择算法
决策树
自体集
computer immune
intrusion detection
negative selection algorithm
decision trees
self-set
