Previous |  Up |  Next


First some new concepts are introduced (extension of the similarity-function, similarity-measure, transitive sets) which facilitate the study of hierarchical classifications. Then an algorithm for forming hierarchical classifications with "good properties" is sketched. Finally, this procedure of classification is described in detail and represented by flowcharts. Theoretical justification of the method is given in Theorems 1 to 4 which show the connection between the properties of transitive sets and hierarchical classifications. Practical use of the method requires a concrete choice of the similarity-function as well as the data given in the corresponding form.
[1] I. C. Lerman: Les bases de la classification automatique. Paris, 1970. MR 0349079 | Zbl 0199.51402
[2] P. Macnaughton, Smith: Some statistical and other numerical techniques for classifying individuals. Home Office Research Unit Report, s 1 - 31. London, 1965.
Partner of
EuDML logo