Skip to main content


Fig. 6 | BMC Medical Informatics and Decision Making

Fig. 6

From: Towards stroke prediction using electronic health records

Fig. 6

Model 3 ROC attained on the validation (a) and test (b) sets, when initialized with weights transferred from models trained with the custom loss function, and then trained to minimize the standard cross-entropy loss. The shaded area displays the 95% confidence interval. The AUC is calculated from the solid line

Back to article page