核聚类算法：聚类是将一组给定的未知类标号的样本分成内在的多个类别，使得同一类中

的样本具有较高的相似度，而不同类中的样本差别大。侧重于软聚类（模糊C-均值——FCM），但其描述手段同样适合于硬聚

类（HCM）等同类问题。-Clustering algorithm: cluster is a group of unknown samples given class label into internal multiple categories, so that the same class

The sample has a high degree of similarity, rather than difference in kind in large samples. Focused on the soft clustering (Fuzzy C- Means FCM), but its descr iption means equally suitable for hard poly

Class (HCM) and other similar issues.