Pelevin S. Models and methods of intelligent support of decision - making based on registry systems

The existent register systems come forward the object of researches: register of agricultural animals and register of cereals. The purpose of dissertation work is an increase efficiency of analysis the accumulated data and decision actual tasks of subject domains which do not decide current possibilities of the inculcated register systems of agricultural animals and cereals, by development and research models and methods of operative analysis sparse, multidimensional data structures and models, methods of intellectual data analysis. In the process researches the methods of planning multidimensional databases, theory of sets and counts, methods system analysis, methods intellectual analysis, methods statistical analysis, probability theory, were used. Scientific novelty: a method is first developed for the increase efficiency of query to sparse multidimensional databases; the method of prognostication critical situations is offered, on the basis of cluster analysis; a multidimensional model and aggregating method of operative analysis is improved; found further development application the methods of estimation possible losses arrived by an enterprise, on the basis of multidimensional data model. Degree of introduction: On the basis of the models and methods offered in a work, modification of register of agricultural animals and register of cereals is made. Test exploitation of software components, on the basis of the offered models and methods, conducted in agricultural enterprises: «Schorsa», Cherkasskaya obl.; «Lan» Cherkasskaya obl. Sphere of the use: work results, except modification of the register systems of agricultural animals and cereals, can be used in the field of development and exploitation of the systems of support of making decision, which are built on the basis of multidimensional depositories of information, in different industries of national economy of Ukraine.