Open AccessUSING ARTIFICIAL NEURAL NETWORKS TO PREDICT POSITIVE DYNAMICS IN THE TREATMENT OF TUBERCULOSIS PATIENT
Author(s) -
M.Z. Ermolitskaya,
Control Processes
Publication year - 2020
Language(s) - English
DOI - 10.22250/isu.2020.66.139-147
Subject(s) - artificial neural network , tuberculosis , mean squared error , studio , dynamics (music) , sample (material) , computer science , artificial intelligence , statistics , medicine , psychology , mathematics , pathology , chromatography , telecommunications , pedagogy , chemistry
R Studio program was used to develop a neural network model that predicts positive dynamics in the treatment of tuberculosis patients in the hospitals. The accuracy of the presented model in the test sample is 99.4%, and the mean square error is 0.013.