Table 3.

Performance of DNN across all data, the training set, and the test set.

Dataset	F1	Precision	Recall	AUROC	AUROC 95% CI	AUROC SE
All data (n = 1349640)	0.931	0.982	0.889	0.738	0.734–0.742	0.002
Training set (n = 1315899)	0.931	0.982	0.889	0.739	0.735–0.743	0.002
Test set (n = 337410)	0.929	0.980	0.886	0.700	0.674–0.727	0.013