最优聚类个数和初始聚类中心点选取算法研究

传统k-means算法的聚类数k值事先无法确定，而且算法是随机性地选取初始聚类中心点，这样容易造成聚类结果不稳定，且准确率较低。本文基于SSE用来选取聚类个数k值，基于聚类中心点所在的周围区域相对比较密集，其次聚类中心点之间距离相对较远的选取原则用来选取初始聚类中心点，避免初始聚类中心点集中在一个小的范围，防止陷入局部最优。试验证明，本文能选取最优的k值，通过用标准的 UCI数据库进行试验，本文采用的算法能选择出唯一的初始中心点，聚类准确率较高，误差平方和较小。

The algorithm research of optimal cluster number and initial cluster center

The cluster k of traditional k - means algorithm could not determine beforehand and the initial clustering centers of k-means algorithm are randomly selected,which may result in low accurary and unstable clustering. In this paper, based on the SSE for selecting the number of clusters k, based on the principle that the clustering center of the surrounding area is relatively dense, and between the clustering center distance is relatively far,to avoid the initial clustering center focused on a small range, prevent fall into local optimum. Tests show that ,this paper can select the optimal value of k, in the case of the number of categories k is given ,used the standard UCI data sets for test. this paper can choose the only center of initial clustering and the higher accuracy,the improved selection of initial centers in this paper have the minimum errors.