A Test for Clusters

Abstract
Tests for the presence of clusters are reviewed. An alternative test is derived that indicates whether data are other than unimodally or uniformly distributed. The test allows for at least 1,000 entities, 20 attributes, and 8 groups. The test is applied to data with known clusters and then to data gathered to solve a problem in lifestyle market segmentation.

This publication has 16 references indexed in Scilit: