Multivariate statistical technique that arranges objects from a data set in clusters, each of which binds objects more similar to one another than those in other clusters.