Application of algorithms of objectifying expert clustering of Multiparameter objects in the analysis of big arrays of information

Abstract

The use of expert assessments is relevant for the development of new complex scientific and technical products, in medical diagnostics, in the study of unusual phenomena. To successfully apply these estimates, it is proposed to use a procedure that allows improving the quality of expert clustering of objects based on the analysis of measured parameters - a procedure that objectifies expert estimates. This interactive procedure is based on elementary assumptions about the properties of objects. The paper describes the features of the implementation of the algorithm of this interactive procedure, proposed by the author. This procedure is essential during the creation of the technology of work to with complex objects that are difficult to fully or even partially formalize. A method for representing the proposed procedure for objectifying expert clustering in terms of the theory of fuzzy sets is described. The developed objectifying procedure was used to create algorithms for medical diagnostics based on pulsed signals of the radial artery. Subjective clustering of the shape of pulse signals was used.