A New Clustering Validity Index based on K-means Algorithm
Xiangru Hou · Journal of Physics Conference Series · 2019
Although cluster analysis has got great achievements, there are many questions in it. In this paper, the question on determining optimal number of clusters in cluster analysis is studied mainly. KMS (K-means Silhouette) for determining optimal number of clusters in K-means clustering algorithm are proposed. KMS (K-means Silhouette) algorithm improves the way of setting initial clustering centers in K-means clustering algorithm, and uses Silhouette validity index to determine optimal number of clusters. The experimental results on artificial datasets indicate the effectiveness of the proposed algorithms.