Comparative Analysis of an Algorithm Based on Random Selection in the Classification of Medical Emblems

Resumen

The article took as an object of study ischemic heart disease (5 classes ,507 objects, 89
characters, including X1- class"strain angina", X2- class "acute myocardial infarction", X3-
class "arrhythmic form", X4- class"cardiosclerosis after infarction", X5- class "permanent
form of atrial fibrillation") in a research work for l = 9 in 103 variants from informative
character complexes were selected, a classification algorithm based on random selection was
developed to assess the classification efficiency of these selected informative character
complexes, and the results were obtained. Also, the effectiveness of the classification of these
informative character complexes was obtained using well-known algorithms on the popular
KNIME analytical platform. Then a comparative analysis of the results obtained from the
developed and existing algorithms was carried out, assessing the effectiveness of the
classification (diagnosis) of selected complexes of informative signs.