Abstract
A method for multidimensional hierarchical clustering tha t is in- variant to monotonic transformations of the distance metri c is pre- sented. The method derives a tree of clusters organized acco rding to the homogeneity of intracluster and interpoint distance s. Higher levels correspond to coarser clusters. At any level the meth od can detect clusters of different densities, shapes and sizes. T he number of clusters and the parameters for clustering are determine d auto- matically and adaptively for a given data set which makes it unsu- pervised and non-parametric. The method is simple, noniterative and requires low computation. Results on various sample data sets are presented.

This publication has 5 references indexed in Scilit: