Cluster Analysis

Abstract

Cluster analysis refers to a family of related techniques that are used to create groups of objects that, in some sense, minimize the distance between the objects within a cluster (Arabie, Hubert, & DeSoete, 1996; Hartigan, 1975). It is generally viewed as a type of exploratory data analysis in which the clusters emerge from the analytical procedures. Cluster analysis has been used extensively since the 1970s by social scientists, market researchers, numerical taxonomists, and many others. Cluster analysis is also at the heart of many current approaches in machine learning, pattern recognition, and data mining.