Kislitsa L. Information technology of making decision support based on time series prognostication in the multicriteria conditions

The reseach object is time series with long and double long memory in various application areas. The research purpose is to encrease the efficiency of multicriteria decision making on the basis of time series prediction with a double long-memory/ The research methods are methods of fuzzy logic techniques, mathematical modeling and system analysis of time series with different nature, the probability theory and mathematical statistics, the methods of multi objective optimization, the theory of computer information systems design. The theoretical results are following: the information technology for multicriteria decision support based on time series prediction with a long and double long-memory, the method for identification of time series with double long memory, that uses a classifier based on fuzzy knowledge base, the mathematical models of time series with double long memory, further development of the hierarchical approach Saati, that uses a tree hierarchy, used time series prediction with double long memory/ The practical results are following: the procedure for double long memory identification, the decision-making algorithm based on time series with double long memory prediction, the algorithmic and software information system decision support with an open architecture. The results of dissertation research have been implemented at Dniester HPP-2, “Kreditprombank ", in Vinnytsa torgovo-promislova Palata, at Company "Spilna sprava" and in the educational process of the Department of automatics and data-measuring engineering at Vinnytsia National Technical University (Ukraine). The application fields are an expert systems, the monitoring and automated management with decision support.