中文说明:K的简单代码意味着带剪影的。此代码作为起点与PCA技术一起提供。绘制数据后,可以找到数据中存在的簇(或不同类)的数量。剪影是一种检查簇数假设的方法,即1个良好的指定,0个观测值可能属于另一个簇,1个错误的指定属于该簇。
English Description:
Straightforward code for K means++ with Silhouette. This code accompanies PCA tehnique, as being a starting point. After ploting the data one can fiind the number of clusters( or different classes) present in data. Silhouette is a methode of checking the assumption of the number of clusters, that is 1 good designation, 0 the observation can be belong to another cluster and -1 a wrong designation to the cluster. The designation to a cluster is done by minimizing distances to cluster centroid.