ABSTRACT

Anomalous pattern A method for separating a cluster that is most distant from any reference point including the grand mean. The method is similar to K-Means at K = 2 except for the never-changing location of the reference point serving as the centroid of the other cluster.