Data analysis for COVID-19 using regression methods

Abstract

With the appearance of the first registered case of corona, as one of the world's most widespread and most dangerous viral infections, the need to monitor and predict the epidemiological situation is growing, both in the world and in our country. In this paper, the epidemiological data of the Republic of Serbia regarding the Corona virus in the period from 2020 to June 2021 are analyzed. Data were analyzed by regression methods, as one of the data mining techniques. Depending on the choice of regression method (simple, multiple and linear), a number of parameters were selected that include the number of persons (positive, tested, deceased, hospitalized and respirator) in relation to the time of the pandemic to make the most accurate prediction. As a result of the research using regression methods, it was found that the trend of development of the Corona virus epidemic is decreasing, i.e. (id est.) that preventive measures as well as the process of vaccination and revaccination have had an effect in the fight against Corona virus.