Fig. 5
From: Using random forest and biomarkers for differentiating COVID-19 and Mycoplasma pneumoniae infections
Precision-recall curve for the random forest classifier. 95% CI, 95% confidence interval; class 1 (COVID-19 positive), AUC = 0.71 (95% CI: 0.45–0.92); class 2 (Mycoplasma pneumoniae positive), AUC = 0.87 (95% CI: 0.75–0.97); class 3 (Co-infected), AUC = 0.40 (95% CI: 0.08–0.73); micro-average ROC curve, AUC = 0.80 (95% CI: 0.69–0.91).
