CLASSIFICATION OF OBJECTS ON THE BASE OF THE EXPECTED INFORMATION VALUE
In this paper applying the expected information value for determining the degree of dissimilarity of structures was proposed. The methodology proposed represents one of numerous attempts at employing the measures defined on the grounds of the information theory for investigating socio-economic phenomena. The expected value of information on transformation of the observed structure into another structure compared with it may be treated as the starting point in the central agglomeration procedure. The paper presents the hierarchic classification by means of full linking of counties in Warminsko-Mazurskie voivodship according to the similarity of the structure of economic entities according to the PKD (Polish Classification of Business Activities) sections so that the possibility of employing the expected information value in the classification procedures was presented.
CEJSH db identifier