Shved A. Models and information technologies of clustering and ranking of group expert's estimates under uncertainty

The object of research: group expert's evaluations, which require structuring and aggregation to form a collective decisions. Aim of the research - explore new approaches and developing mathematical models of clustering and ranking group expert's assessments based on current theories, and information technologies for analysis of expert's estimates. Methods of research: methodology of systems analysis, decision theory, the theory of evidence, the theory of plausible and paradoxical reasoning, adaptive robust procedures. Theoretical, practical results and novelty: the complex method of expert's estimates clustering based on the integrated use of adaptive robust estimates is proposed and tested, the decision rules of clustering are generated; the mathematical models of expert's judgments and the mathematical models of expert's estimates ranking under multicriteria, multialternative, individual expert uncertainty and highly conflicting sources of expert information are further developed; the complex method of building a final ranking of expert's estimates based on the elements of mathematical theory of evidence and using a set of conjunctive rules of combination, and taking into account the distance measure of individual evidence to determine the order of combination is improved. The results of the research implemented in the educational process of the Petro Mohyla Black Sea State University, National University of Shipbuilding, Mykolaiv National University of V.A. Sukhomlinsky and , as well as in the activities of SP "GRASKO" and PB "TEHNOLIS". Application areas: decision support systems and expert systems.