Research in machine learning methods for solving problems of the medical profile

Abstract

The work is devoted to the study of machine learning methods for solving medical problems. The aim of the work is to analyze machine learning methods to improve the accuracy and reduce the time for diagnosing diseases of the genitourinary system in children. The object of research is machine learning methods. The subject of the study is a classifier of diseases of the genitourinary system of patients of the Dnipropetrovsk Re-gional Children's Clinical Hospital "Dnepropetrovsk Regional Council". As a result of the study, the following tasks were solved: an analysis of the literature on the applica-tion of machine learning methods to diseases of the genitourinary system was made; a program was developed to extract the necessary information on statements in a semi-automatic mode; Python libraries and part of machine learning methods were analyzed; primary analysis and processing of data was carried out; applied methods of classifica-tion, feature selection and filling in missing values; the obtained results were analyzed and the substantiation of the research results in the subject area was made.