Methods of intellectual data analysis in COVID-19 research

The paper presents an original method for solving the problem of finding a connection between course epidemic and socio-economic, demographic climatic factors. was applied to solve this 110 countries world using set corresponding curves COVID-19 growth rate period from January 2020 August 2021. Hierarchical agglomerative clustering applied. Four large clusters with uniform were identified – 11, 39, 17 13 countries, respectively. Another 30 not included in any cluster. Using machine learning methods, we differences geographical indicators selected world. most important by which differ each other are amplitude temperatures throughout year, high-tech exports, Gini coefficient, size urban population general population, index net barter terms trade, growth, average temperature, territory (land area), number deaths due natural disasters, birth rate, coastline length, oil reserves, agglomerations more than 1 million etc. This approach (the use combination classification methods logical-statistical analysis) has been used anyone before. found patterns will make it possible accurately predict epidemiological process belonging different clusters. Supplementing autoregressive models automate forecast improve its accuracy.