基于模糊邻近关系的结构聚类

在有序粒度空间理论的基础上,提出了基于模糊邻近关系的结构聚类分析理论和方法.首先,给出了依据距离的一致聚类的概念,提出了模糊粒度空间的一致聚类特征;其次,给出了模糊邻近关系结构聚类的粒度表示,并获得了基于模糊邻近关系结构聚类的快速算法;进而,提出了获取最佳聚类的新方法,并且这一方法是全局最优的;最后,给出了通过两个模糊邻近关系的交运算获取结构聚类融合的方法.为复杂系统结构分析研究提供了一整套理论工具和方法.

Structural clusters based on fuzzy proximity relations

On the basis of fuzzy granular space, structural clustering analysis based on fuzzy proximity relations is proposed. Firstly, the concept of consistent clustering according to metric is proposed, and the research on consistent clustering of a fuzzy granular space is given. Secondly, the granular representation on structural clustering of a fuzzy proximity relation is studied, and the algorithm to obtain its structural clustering is given. Thirdly, the research to determine the optimal clustering based on a fuzzy granular space from a fuzzy proximity relation is proposed, and a new method to obtain it is given, which the new method is overall. Finally, the fusion technology is attained by applying the intersection operation of two fuzzy proximity relations. These conclusions provide a comprehensive theoretical tool and methodologies for studying structural analysis of complicated systems.