I. Introduction
Using AI algorithms in medicine opens up huge opportunities for the different domain of healthcare. As a consequence, todays, various algorithms of AI are widely used in the direction of disease detection, interpretation and segmentation of medical images as well as classification of diseases [1 –3, 15 –18]. But in healthcare, there are cases when it is more important to determine the factors (drug, food, physical activity etc.) affecting the patient’s condition and to determine the degree of influence of these factors than to diagnose the disease. In particular, the monitoring of chronic diseases is one of this crucial case. Because, in the process of daily monitoring of the patient, it is necessary to determine what factors and to what extent cause its condition to improve or to worsens. To solve such kind of problem, machine learning methods of evaluating the importance of features in the dataset has been used.