Registration Number
0224U001475,
0120U100025
, R & D reports
Title
Methods and means of processing, consolidating and analyzing personalized medical information
popup.stage_title
Head
Melnykova Nataliya I., Кандидат технічних наук
Registration Date
23-01-2024
Organization
Lviv Polytechnic National University
popup.description2
A model of displaying the patient's condition in a multidimensional space of conditions was developed, which, due to additional measurements and parameters, ensured an increase in the accuracy of forecasting; a method of two-stage processing of small data sets based on a hierarchical predictor, which, due to clustering and forecasting, ensures an increase in the accuracy of processing new input data sets; a method of grouping machine learning models, which, taking into account the main features and repeated training, ensures an increase in the accuracy of data prediction; the method of reducing the dimensionality of the input data, which, thanks to classifiers, associative rules and the generalized rank of features based on the Jacquard index, ensures an increase in the accuracy of selecting priority features on large data sets. The method of filling data gaps has been improved, which, using associative and production rules, ensures the improvement of the model's resistance to data errors; the method of finding a change in the patient's condition, which, using the space of conditions and the analysis of the characteristics of the condition, ensures an increase in the accuracy of the selection of the treatment scheme; the method of multimodal data consolidation, which provides aggregation of data with different structures due to the preliminary determination of the structure and semantics of the data. The theory of support for personalized decision-making was further developed, which is based on the development of a model for displaying the patient's condition in a multidimensional space of conditions and the developed methods of two-stage processing of small data sets, grouping of machine learning models, reducing the dimensionality of input data, filling data gaps, searching for changes in the patient's condition and consolidation multimodal data. The developed methods were used to develop software modules of information systems for decision-making support.
Product Description
popup.authors
Basystiuk Oleh
Boyko Natalya I
Brusencov Heorhii
Vyklyuk Yaroslav I.
Vovk Olena B
Zavaliy Taras
Zagorodniy Ivan
Paterega Yurii
Pobereyko Petro
Skopvskiy Stepan
Fedushko Solomiia S.
Tsap Vladyslav
Shakhovska Natalya Bogdanivna
popup.nrat_date
2024-01-23