摘要
提出一种基于免疫原理的动态聚类算法,它能在噪声环境下得到任意形状的聚类,并能有效地实现动态聚类操作.算法包括3个步骤:首先基于生物免疫机制得到一个反映当前数据分布特征的抗体集合;然后使用最小生成树方法得到聚类的初始结构;最后针对数据库的更新设计了动态聚类算法.仿真结果表明了该算法实现动态聚类的有效性.
A dynamic clustering algorithm based on immune principle is proposed. Arbitrary shape clusters are generated in the presence of noise by using the algorithm, and dynamic clustering is implemented efficiently. A set of antibodies are obtained based on biology immune mechanism reflecting the distributing information of current data set. The initial clustering structure is constructed by using the minimum spanning tree method. The dynamic clustering algorithm is developed to update the clusters. Experiment results show that the effectiveness of the proposed algorithm.
出处
《控制与决策》
EI
CSCD
北大核心
2007年第4期469-472,共4页
Control and Decision
关键词
免疫原理
动态聚类
聚类
Immune principle
Dynamic clustering
Clustering
