MONITORING AND MODELING OF THE EPIDEMIOLOGICAL SITUATION USING DATA MINING

Abstract

In this study, based on data mining technology, machine learning methods and cluster analysis, the identification of tasks is corrected, numerical solution algorithms for a mathematical model for the prevalence of socially significant HIV infection in Kazakhstan are described.

Data mining technology in modeling the HIV situation is particularly relevant, as it is used to develop maps of short-term prognosis of morbidity in Kazakhstan and the regions of the country. The study was conducted from statistically taken data on the prevalence of HIV infection in Kazakhstan over the past 10 years (2010-2020). Information technologies, including Data mining technologies, allowed the authors to describe disease schedules, identify risks, check statistical forecasts of diseases.        The main part of the study describes such indicators as the numerical solution algorithm and the construction of a mathematical model of HIV epidemiology by classifying regions into homogeneous groups. Data Mining classification methods were used to process HIV-infected people and analyze their condition in the region. The prognosis of the morbidity of the population of Kazakhstan was carried out using the Statistica application software package. An effective numerical solution algorithm for mathematical modeling allows testing developments on real data.