Fig. 1From: Predicting COVID-19 mortality risk in Toronto, Canada: a comparison of tree-based and regression-based machine learning methodsAUC (left panels) and Brier’s score (right panels) over the 200 testing samples of the repeated split sample validation for the methods including all the considered predictors (top panels) and excluding hospital use variables (ever hospitalized, ever ICU use and ever intubated) for COVID-19 as predictors (bottom panels)Back to article page