Table 4.

Performances of logistic regression and deep neural network models for MI and stroke.

Variable	MI				Stroke
	Missing Value (+)		Missing Value (−)		Missing Value (+)		Missing Value (−)
	LR	DNN	LR	DNN	LR	DNN	LR	DNN
Accuracy	0.686	0.911	0.727	0.917	0.681	0.883	0.682	0.897
F1 score	0.696	0.908	0.734	0.916	0.687	0.874	0.690	0.889
Precision	0.675	0.939	0.715	0.927	0.674	0.935	0.674	0.956
Recall	0.718	0.879	0.754	0.907	0.700	0.824	0.707	0.832
AUC	0.686	0.968	0.727	0.972	0.681	0.948	0.682	0.951

AUC: Area under the curve, DNN: deep neural network, LR: logistic regression, MI: myocardial infarction. p < 0.0001 for all two groups.