Partitional and hierarchical data clustering methods

Katarzyna Stąpor

Abstract


In this paper the most important partitional and hierarchical data clustering algorithms arę described. Among partitional algorithms those based on point and linear prototypes arę discussed. In hierarchical algorithms different similarity measures arę described. Also, neural network clustering based on Kohonen network is described.

Keywords


clustering; partition matrix; dendrogram; similarity measure



DOI: http://dx.doi.org/10.21936/si2004_v25.n1.636