CLASSIFICATION OF DATA IN THE DECISION-MAKING SYSTEM FOR THE PREVENTION OF TYPE 2 DIABETES BASED ON THE METHODS OF MACHINE LEARNING

Abstract

According to the statistics of the World Health Organization, more than one-third of the adult population have symptoms of developing type 2 intellectual data analysis of diabetes mellitus (T2DM). Scientific advances in information and telecommunication technologies, the development of IoT technologies (Intertnet of Things) and the widespread use of artificial in- telligence (AI) methods provide tremendous opportunities for the introduction of telemedicine in the field of health.

The article presents the experimental results of the medical data classification for the ear- ly detection and prevention of the occurrence possibility of T2DM in the decision making sys- tem, using the methods of machine learning and intellectual data analysis. The article also ex- amines the architecture of the smart medical decision making system for the prevention of T2DM disease, which is designed to identify individuals at a risk of a possible development of the disease (prediabetes).

The article compares the efficiency criteria of the considered AI methods for the preven- tion of T2DM disease and evaluates the performance of the following methods of machine learning: Support Vector Machine (SVM), Neural Networks (NNs), Logistic Regression (LR), Naive Bayes (NBs), Random Forest (RF), Decision Trees (DTs) and k-Nearest Neighbors (K- NNs). Various performance indicators of methods are considered: Accuracy, F-score, Preci- sion, Recall and ROC Area. An important aspect of the study is the use of various performance criteria to evaluate machine learning methods for early detection and prevention of T2DM. The study proves the feasibility of obtaining high accuracy in the analysis of medical data by the methods SVM, LR, NNs and RF.