Graph Based Visual Assessment Cluster Tendency for Unlabeled Data Sets

Download Now
Provided by: Binary Information Press
Topic: Data Management
Format: PDF
Visual methods have been broadly deliberated and performed in cluster data analysis. Enhanced-Visual Assessment Tendency (E-VAT) algorithm normally represent D as an nxn image where the objects are reordered to represent the concealed cluster structure as dark blocks along the diagonal of the image. A major limitation of the E-VAT is the lack of ability to highlight cluster structure when D contains complex shaped datasets. The proposed Graph based Visual Assessment Cluster Tendency (GVACT) system is performed for generating Reordered Dissimilarity Image (RDI) that combines E-VAT for reordering with weighted graph analysis of pairwise data, local scaling parameter, combinations of graph Laplacian and obtaining a better graph embedding in the present paper.
Download Now

Find By Topic