Study on determining the optimal number of cluster in a dataset

摘要

Clustering analysis grouping of the data object that depend upon the information search in the data that describe the object and their relationship. It split the dataset into groups of the data object within the same cluster are more matching to the object in the different clusters. The objective of this article to find out the “good” number of clusters. It is of three types of categories: internal measures, external measures and cluster validation.