Table 2.

Performance comparison of four algorithms on training and testing datasets.

Machine learning models	Training accuracy	Training sensitivity	Training specificity	Training F1 score	Training AUROC	Testing accuracy	Testing sensitivity	Testing specificity	Testing F1 score	Testing AUROC
Logistic regression	0.859	0.863	0.855	0.870	0.904	0.809	0.831	0.781	0.827	0.852
Naive bayes	0.787	0.897	0.655	0.822	0.897	0.691	0.843	0.507	0.750	0.826
XGBoost	0.867	0.832	0.909	0.873	0.927	0.796	0.764	0.836	0.805	0.852
Multilayer perceptron	0.890	0.858	0.929	0.895	0.906	0.840	0.820	0.863	0.849	0.898