A New Method to Determine Cluster Number Without Clustering for Every K Based on Ratio of Variance to Range in K-Means

In many clustering algorithms such as K-means and FCM, the cluster number K needs to be known beforehand. this paper, we propose a new method determine without for every in K-means. We introduce statistics RVR (ratio of variance range) conduct Monte Carlo analysis its characteristics. Based on RVR, an algorithm perform utilizing it. evaluate effectiveness by performing simulation test with different types datasets; first, real datasets, whose clusters components are second, synthetic datasets. observe significant improvement speed quality determining therefore clustering. Finally, hope proposed used efficiently widely multidimensional data.