The Research Of Clustering Algorithm And Its Application Based On Dynamic Fuzzy Relations

At present the research of clustering analysis algorithm for the dynamic fuzzy data is still comparatively rare. Therefore, in this paper,based on the dynamic fuzzy logic theory and combined with the theory of traditional clustering analysis and fuzzy clustering analysis, the author tries to do some research in clustering analysis algorithm for dynamic fuzzy data, here's the main work:1.Studied several types of clustering analysis algorithm for dynamic fuzzy data :(1) Transitive closure clustering algorithm based on dynamic fuzzy equivalence relation .(2) Maximal tree clustering algorithm based on dynamic fuzzy similarity relation。(3)Clustering algorithm based on dynamic fuzzy similarity boolean matrix.2.Expounded the advantages and disadvantages of various algorithms. Transitive closure clustering algorithm based on dynamic fuzzy equivalence relation and maximal tree clustering algorithm based on dynamic fuzzy similarity relation are compared. The two equivalence of cluster analysis algorithm is proved with theory.3.For the distortion of transitive closure clustering algorithm, the article gives clustering analysis algorithm based on the optimal dynamic fuzzy equivalent matrix.Through this study, the research content of dynamic fuzzy logic is further enriched and the scope of cluster analysis is broadened. In addition, this article provides some methods and guidance of the clustering analysis for the dynamic fuzzy data. However, related work is just a preliminary attempt. A lot of further work needs to be explored.