基于免疫网络的无监督式分类算法

An immune network based unsupervised classifier

基于免疫网络原理,提出了一种新的无监督式分类算法。首先基于形态空间理论给出了抗体、抗原和免疫网络的形式化定义,建立了抗体克隆选择、高频变异以及免疫记忆的动态模型和相应的数学方程,最后给出了分类过程。实验表明该算法的分类精度要高于其它传统的聚类算法,并具有很好的持续学习、动态调节、特性记忆等特性。如果把抗体视为某种既定模式,合理地调整抗原集合,则该模型具有广泛的用途。

A novel unsupervised classification algorithm based immune network was presented.First,the formal definitions of antibodies,antigens and immune network were given according to shape space theory.Second,mathematical models and corresponding equations were established,such as the clonal selection and high-frequency mutation of antibodies,and the immunological memory.Finally,the process of unsupervised classification was presented.Experimental results showed that the algorithm achieved higher classification accuracy than other traditional clustering algorithms,and had some better characters such as continuous learning,dynamic adjustment,and remembering features.If the antibody is regarded as a given model,and rationalizes the antigens collection,then the model has a wide range of applications.