From: Forecasting patient flows with pandemic induced concept drift using explainable machine learning
 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | Overall |
---|---|---|---|---|---|---|---|
Prophet | 0.774 | 0.76 | 0.591 | 1.584 | 1.064 | 1.133 | 1.045 |
Stacking | 0.725 | 0.693 | 0.56 | 1.167 | 0.888 | 0.982 | 0.855 |
Random Forest | 0.743 | 0.73 | 0.631 | 1.026 | 0.962 | 0.969 | 0.845 |
Averaging | 0.703 | 0.689 | 0.566 | 1.052 | 0.871 | 0.967 | 0.815 |
Gradient Boosting | 0.736 | 0.732 | 0.626 | 0.894 | 0.916 | 1.043 | 0.810 |
Voting | 0.692 | 0.684 | 0.558 | 1.009 | 0.886 | 0.947 | 0.800 |
CatBoost | 0.754 | 0.737 | 0.63 | 0.836 | 0.872 | 1.069 | 0.796 |