Abstract
The COVID-19 pandemic has emerged as the world’s most serious health
crisis, affecting millions of people all over the world. The majority of
nations have imposed nationwide curfews and reduced economic activity to
combat the spread of this infectious disease. Governments are monitoring
the situation and making critical decisions based on the daily number of
new cases and deaths reported. Therefore, this study aims to predict the
daily new deaths using four tree-based ensemble models i.e., Gradient
Tree Boosting (GB), Random Forest (RF), Extreme Gradient Boosting
(XGBoost), and Voting Regressor (VR) for the three most affected
countries, which are the United States, Brazil, and India. The results
showed that VR outperformed other models in predicting daily new deaths
for all three countries. The predictions of daily new deaths made using
VR for Brazil and India are very close to the actual new deaths, whereas
the prediction of daily new deaths for the United States still needs to
be improved.