Abstract: Clustering is a very important topic in the field of data mining. An adaptive artificial immune algorithm for clustering, based on aiNet network model, is presented in this paper. The algorithm has the ability to achieve final network structure well-imaging the crude data feature and wide application situation because of its immune suppression among the antibodies which suppression threshold is self-adaptive redressal to the whole network structure. Experiment results demonstrate that the algorithm is capable of gaining excellent and stable clustering result, moreover, the more important thing is that the member of control parameter in this algorithm is less than that in other control algorithm.

Key words: Data mining, artificial immune system, clustering analysis, adaptability

摘要： 聚类是数据挖掘领域中的一个重要研究课题。基于人工免疫网络aiNet，本文提出一种自适应的人工免疫聚类算法，采用网络抗体间抑制阀值随网络进化而自适应变化的淘汰策略，使最终网络结构更符合原始数据的内在结构，也使算法具有较好的泛化能力。通过仿真实验，该算法表现出很好的稳定性和良好的聚类效果。同时，需要用户设置的参数很少。

关键词: 数据挖掘, 人工免疫系统, 聚类分析, 自适应