期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2015
卷号：6
期号：3
页码：210-213
语种：English
出版社：Ayushmaan Technologies
摘要：Affinity propagation clustering be an exemplar-based scheme that comprehends via the handover of each data point to its nearest exemplar, where exemplars be acknowledged viapassing messages on bipartite graph. There are two kinds of messages passing on bipartite graph. They are responsibility and availability, collectively called ’affinity’. The goal of this paper is to propose a dynamic variant of AP clustering, which can accomplish equivalent clustering performance with traditional AP clustering by just middle with the current clustering results according to new arriving objects, slightly than re-implemented AP clustering on the whole dataset. Therefore, a great deal of time can be saved, which makes AP clustering well-organized, sufficient to be used in dynamic environment.
