基于决策树及遗传算法的人工免疫入侵检测算法

Intrusion Detection Algorithm of Artificial Immune Based on Decision Tree and Genetic Algorithm

针对传统的否定选择算法中生成大量无效抗体,并且抗体之间缺乏多样性的问题,设计了基于决策树及遗传算法的人工免疫入侵检测算法。将决策树和遗传算法引入传统的否定选择算法中,利用决策树计算抗原和抗体之间的亲和力,提出了新fitness的计算公式,并利用抗体浓度衡量抗体集的多样性,将低浓度抗体代替高浓度抗体,实现了抗体的多样性,确保了当抗体集的数量一定时尽可能覆盖最大的非自体集空间,以提高抗体集的性能。

Aiming at solving the problem that there were large amounts of ineffective antibodies and the antibodies were lack of diversity in the traditional negative selection algorithm ,this paper designed intrusion detection algorithm of artificial immune based on decision tree and genetic algorithm. The decision tree and the genetic algorithm were introduced into the traditional negative selection algorithm, the affinity between antibody and antigen was calculated using decision tree, the new formula of fitness was raised. The diversity of antibody set was measured by concentration of antibody, and the high concentration antibodies were replaced by the low concentration antibodies to achieve the diversity of the antibody set. When the quantity of the antibody set was kept at a constant, the nonself set space could be covered as large as possible So as to enhance the capability of the antibody set.