Short-term forecasting COVID-19 cumulative confirmed cases: Perspectives for Brazil
Short-term forecasting COVID-19 cumulative confirmed cases: Perspectives for Brazil
Matheus Ribeiro,Ramon Gomes da Silva,V. Mariani,L. Coelho
TLDR
Regression models are employed to forecasting COVID-19 cases in the Brazilian context and SVR and stacking ensemble are the most suitable tools to forecasting COVID-19 cases in the evaluated scenarios.
Abstract
Highlights • Regression models are employed to forecasting COVID-19 cases in the Brazilian context.• Data from ten states with a high incidence of COVID-19 are adopted.• Models for multi-step-ahead forecasting are evaluated.• Out-of-sample forecasting errors lower than 6.9% are achieved by best models.• SVR and stacking ensemble are the most suitable tools to forecasting COVID-19 cases in the evaluated scenarios.