The object of research is the technology of processing, management and intellectual analysis of medical data in automated systems.
The goal of the work is to increase the efficiency of processing and analysis of heterogeneous data with missing values in automated medical information processing systems.
Features of medical data sets, requirements for collection, processing and analysis of medical data, data processing methods are determined. It is established that: medical data are heterogeneous, ie those that include missing values, inhomogeneous, shifted, unbalanced data. Such data properties lead to erroneous and erroneous results when using predictive analytics to diagnose certain conditions of the patient.
A model and method of missing data management have been developed, which, in contrast to the known ones, take into account qualitative and quantitative assessment of missing values, assessment of data absence mechanism, quantitative assessment of input data and assessment of efficiency of missing data processing method.
A method for identifying interesting rules based on associative data analysis has been developed, which, unlike the known ones, takes into account objective and subjective assessments of determining the interest of rules, which allows to reduce the number of rules while preserving the value of knowledge. The use of the developed technology allows to reduce the number of interesting rules by 44.83%.
Head: Skarga-Bandurova Inna S.. Research on Methods of Date Analysis in Medicine. (popup.stage: ). Volodymyr Dahl East Ukrainian National University. № 0220U102757