| Summary: | Evaluation of clustering results (or cluster validation) is an important and necessary step in cluster analysis, but it is often time-consuming and complicated work. We present a visual cluster validation tool, the Cluster Validity Analysis Platform (CVAP), to facilitate cluster validation. The CVAP provides necessary methods (e.g., many validity indices, several clustering algorithms and procedures) and an analysis environment for clustering, evaluation of clustering results, estimation of the number of clusters, and performance comparison among different clustering algorithms. It can help users accomplish their clustering tasks faster and easier and help achieve good clustering quality when there is little prior knowledge about the cluster structure of a data set.
|
|---|