Abstract
Abstract
In March 2020, rapidly spreading across the world, the severe acute respiratory syndrome coronavirus 2 reached Poland. Since then, many efforts have been made to develop methods to forecast the coronavirus disease-2019 (COVID-19) pandemic spread and to prevent its negative consequences. In this paper, we presented one of such methods, a simplified way of building a data-driven model for predicting the daily number of new coronavirus infections.
Our method is based on parameter selection of the exponentially modified Gaussian cumulative curve, where the obtained curve should describe the curve of a total of COVID-19 cases in Poland with the best possible fit.
We showed that a simplified modelling approach can give good correlations between model values and actual COVID-19 cases data. By forecasting during the COVID-19 epidemic in Poland, we obtained a high enough accuracy for our model to be considered a valuable and helpful tool for making health policy.
Funder
Akademia Górniczo-Hutnicza im. Stanislawa Staszica
Publisher
Cambridge University Press (CUP)
Subject
Infectious Diseases,Epidemiology
Reference22 articles.
1. A contribution to the mathematical theory of epidemics;Kermack;Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences,1927
2. 9. Interdisciplinary Modeling Center of the University of Warsaw (2020) COVID-19 ICM Epidemiological Model. https://covid-19.icm.edu.pl/en/model-description/ (Accessed 22 January 2021).
3. Universality in the distance between two teams in a football tournament
4. Comparing SARS-CoV-2 with SARS-CoV and influenza pandemics
5. 10. StatSoft Polska (2020) Statistical analysis of coronavirus cases in Poland. https://www.statsoft.pl/systemy-data-mining/statystyczna-analiza-zachorowan-na-koronawirusa-w-polsce/ (Accessed 22 January 2021).