scientific article; zbMATH DE number 3936255
Let \(X=\{x^ 1,...,x^ n\}\) be a set of elements to be classified. As has been shown by the author, ibid. 19, No.4, 55-60 (1984; Zbl 0564.62050), the rigorous construction of a total hierarchical classification requires an ultrametric distance on X, or, equivalently, a scatter function. By fundamental algorithms for hierarchical classification we present methods for the construction of hierarchical classifications based on one of these two elements. The algorithm CLAS.F.1, based on an ultrametric distance is, in fact, \textit{S. C. Johnson}'s algorithm [Psychometrika 32, 241--254 (1967; Zbl 1367.62191)] somewhat modified in order to eliminate the existing ambiguities of the original version and to assure the uniqueness of the construction. The algorithm CLAS.F.2 requires the scatter function and it is shown that the hierarchy constructed by this algorithm, corresponding to a scatter function \(\gamma\), is the same as the hierarchy constructed by CLAS.F.1, if an ultrametric compatible with \(\gamma\) is used.
- scientific article; zbMATH DE number 3899973 (Why is no real title available?)
- scientific article; zbMATH DE number 4020249 (Why is no real title available?)
- scientific article; zbMATH DE number 4123085 (Why is no real title available?)
- scientific article; zbMATH DE number 475231 (Why is no real title available?)
- scientific article; zbMATH DE number 2156180 (Why is no real title available?)
- Parallel Processing and Applied Mathematics
- Nesting algorithm for multi-classification problems
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3707149)