Classic: A hierarchical clustering algorithm based on asymmetric similarities (Q801622)

This paper presents a new and interesting hierarchical clustering algorithm called CLASSIC, based on an iteratively defined nested sequence of the nearest neighbors relation (NNR) which is defined in terms of a given asymmetric matrix of the similarities of data items. Application of this new approach is shown in automatic detection of gestalt clusters, and a method is introduced for obtaining asymmetric similarities directly from the Euclidean distances since almost all existing clustering techniques have been based on a symmetric measure.