Forecast and study of time series dependencies on the example of an increase in the incidence of COVID-19

A. A. Butov, A. S. Shabalin

Ulyanovsk State University, Ulyanovsk, Russia

Abstract: The paper analyzes the dependence of the number of new cases of the SARSCoV-2 virus on the values in previous days. Significant lags for incidence in Russia were identified and an attempt has been made to explain them. Correlograms for different countries are compared. For predict-ing the number of new cases of the disease, a linear regression model is presented as well as the quality criteria that determine this model.

Keywords: time series, autocorrelation function, machine learning, linear regression, covid-19.

UDC: УДК 004.942

Received: 06.04.2022
Revised: 13.05.2022

Bibliographic databases:

•