论文部分内容阅读
本文提出了一种新的高效软聚类方法──截集模糊Ｃ－均值聚类（Ｓ２ＦＣＭ）．该方法将传统的硬聚类（ＨＣＭ）和经典的模糊聚类（ＦＣＭ）相统一，并进行了合理的推广而得出的一种有效的聚类方法，Ｓ２ＦＣＭ聚类算法的思想更符合人的分类习惯，模拟实验表明，其总体性能优于ＦＣＭ算法和ＨＣＭ算法。
In this paper, we propose a new efficient clustering method-cutoff fuzzy C-means clustering (S2FCM). The method combines the traditional hard clustering (HCM) with the classical fuzzy clustering (FCM) and makes an effective clustering method which is reasonably generalized. The idea of S2FCM clustering algorithm is more in line with human The classification experiment shows that its overall performance is better than FCM and HCM algorithms.