Article
Summary:
When processing automatically experimental data the need often occurs to replace a certain measured quantity by one or more properties (i.e. the need of binarization of the given quantily).
In the present paper a method is described for the automatic determination of the best binarization of a quantity with respect to the given properties on the basis of the given experimental material. When the research worker determines the maximal number of properties that the quantity is to be devided into and the properties with respect to which the suitability is to be measured, the computer gives all the optimal binarizations.
References:
[1] L. A. Goodman W. H. Kruskal: Measures of Association for Cross Clasifications. JASA, 49 (1954).
[2] P. Hájek I. Havel M. Chytil:
GUHA-metoda systematického vyhledávání hypotéz. Kybernetika 2, 1966, 1, 31-47.
MR 0202606
[3] P. Hájek I. Havel M. Chytil: GUHA-metoda systematického vyhledávání hypotéz II. Kybernetika 3, 1967, 5, 430-437.
[4] P. Hájek: Problém obecného pojetí metody GUHA. Kybernetika - v tisku.