作　　者： ; ;

机构地区： 西安电子科技大学电子工程学院电子工程系

出　　处： 《模糊系统与数学》 1992年第2期77-85,共9页

摘　　要： 模糊c均值(FCM)聚类算法在模式识别领域中得到了广泛的应用,但FCM算法在大数据集的情况下需要大量的CPU时间,令用户感到十分不便,提高算法的速度是一个急待解决的问题。本文提出的双层FCM聚类算法是一种快速算法,它体现了硬聚类和模糊聚类的结合,以硬聚类的结果对模糊聚类的初始值进行指导,从而明显地缩短了迭代过程。双层FCM算法所用的CPU时间仅为FCM算法的十三分之一,因而具有很强的实用价值。 The fuzzy c-means (FCM) clustering algorithms are widely applied in pattern recognition. However, one of the most frequent operational complaints about FCM is that it may consume-for large data sets-large amounts of CPU time. There is sufficient incentive to accelerate computation time in FCM. This paper presents a fast FCM algorithm with two layers. It is amergence of hard clustering and fuzzy clustering. The result of hard clustering is used to initialize the c cluster centers in fuzzy clustering, and then the number of iteration steps is reduced. The CPU time of the FCM with two layers is reduced to about one thirteenth of the time required forFCM implementation, therefore, our new algorithm can be applied for practical purposes.

关 键 词： 模糊聚类 模糊集 模式识别 硬聚类 算法

领　　域： [理学] [理学] [理学] [理学]