A Data Clustering Tool with Cluster Validity Indices
Haiyan Qiao, Brandon Edwards · 2009
Data clustering is an important procedure to detect hidden patterns of a data set in a variety of fields, yet clustering analysis is a challenging problem, because many factors play together in devising and selecting a well tuned clustering technique and there are no predefined classes or examples to show whether the clusters are valid or not. In this paper, cluster validation methods are reviewed, and an extended tool with validation indices is developed.