/NowcastForecastCovid

Primary LanguageMATLABGNU General Public License v3.0GPL-3.0

Nowcasting and Forecasting COVID-19: Accuracy and the Stochastic Nature of Transmission

All the scripts listed here are written in MATLAB and are part of a project that uses a susceptible-exposed-infected-removed-type (SEIR-type) model to calibrate the coronavirus disease 2019 (COVID-19) dynamics and provide forecasted scenarios using different techniques. The project members are Vinicius V.L. Albani, Roseane A.S. Albani, Eduardo Massad, and Jorge P. Zubelli.

#Contents

There is only one main folder named "Codes". The scripts in the folder "Codes" estimate a SEIR-type model from daily reports of COVID-19 in different places and provide forecasted scenarios using different techniques. The script “Running_20220328.m” runs all the scripts used to calibrate the model, to provide forecasts, and to access the quality of the predictions using out-of-sample data (backtest). The codes with Boot refer to bootstrap, and provide the results for the example of the four waves of infections in New York City.

Contact: Prof. Vinicius V.L. Albani: v.albani@ufsc.br.