Melnykova N. Models and methods for supporting personalized solutions in medical systems.

The dissertation solved an important scientific and applied problem of developing and improving models, methods and tools of machine learning in the tasks of classification, clustering, forecasting and visualization of the results of personal data processing for adapting medical solutions to the patient. The work introduces the concepts of personalization, personalized solutions, and patient medical data. Legal requirements for the use of personalized medical data, existing practical solutions reflecting the main approaches to the use of medical data and taking into account personalization based on the analysis of existing artificial intelligence systems in the field of medicine are analyzed. A comparative analysis of classical methods was carried out and the limitations characteristic of the processing of personalized medical data were determined. The biggest limitations during the processing of large and small multimodal medical data sets by universal methods were determined, which identified an actual scientific and applied problem regarding the development of new or improvement of existing methods of artificial intelligence to increase the accuracy of the process of processing large and small multimodal medical data to find personalized solutions and form tasks scientific research.