iVAT and aVAT: Enhanced Visual Analysis for Cluster Tendency Assessment

Given a pairwise dissimilarity matrix D of a set of n objects, visual methods (such as VAT) for cluster tendency assessment generally represent D as an n × n image I(D̃) where the objects are reordered to reveal hidden cluster structure as dark blocks along the diagonal of the image. A major limitation of such methods is the inability to highlight cluster structure in I(D̃) when D contains highly complex clusters. To address this problem, this paper proposes an improved VAT (iVAT) method by combining a path-based distance transform with VAT. In addition, an automated VAT (aVAT) method is also proposed to automatically determine the number of clusters from I(D̃). Experimental results on several synthetic and real-world data sets have demonstrated the effectiveness of our methods.