Search academic texts

Information × Registration Number 0221U105560, 0119U101283 , R & D reports Title To develop distributed methods for analyzing text information and effective procedures for recognizing complex structural objects popup.stage_title Head Rаspopov Viktor B., Кандидат фізико-математичних наук Registration Date 22-07-2021 Organization Research and educational center of applied informatics NASU popup.description2  Two non-equivalent transformations of the labeling scheme ("a coloring of labelings of a column of a schema " and "a transformation of a labeling scheme with respect to the selected subset of objects") were introduced and investigated. A method is developed for applying the operation of coloring the labelings of a schematic column to transform a labeling scheme into a quasi-equivalent scheme, in which it is possible to perform uncoupling operations with any column of the schema. This property of the operation opens the way to create a method for solving the problem of "grammar recovery", i.e. a construction of the labeling scheme, which specifies a given set of mappings based on equivalent transformations of the labeling scheme. The Bayesian approach was used for introducing a method for determining risk groups for the development of certain complications that accompany COVID-19 disease, based on the results of determining point mutations in the genome of patients. Applying this methodology would radically improve the strategy to combat the COVID-19 pandemic. A method for improving the differential diagnosis of pathologies associated with aggregation of blood cells was suggested. It uses a recognition algorithm based on the Bayesian procedure. Product Description popup.authors Brovarnyk Vyacheslav V. Vahis Oleksandra A Гупал А.М. Tkachov Igor I. popup.nrat_date 2021-07-22 Close