Table 1.

Performance of five algorithms with two feature groups.

Algorithm	Feature	AUC	95%CI	Accuracy	Sensitivity	Specificity
RF	Group A	0.897	0.894–0.900	0.824	0.518	0.934
	Group B	0.952	0.950–0.954	0.908	0.733	0.971
KNN	Group A	0.584	0.579–0.588	0.759	0.227	0.950
	Group B	0.881	0.879–0.883	0.873	0.572	0.980
XGB	Group A	0.892	0.889–0.895	0.839	0.596	0.926
	Group B	0.950	0.948–0.952	0.900	0.757	0.952
SVM	Group A	0.925	0.922–0.928	0.860	0.683	0.923
	Group B	0.969	0.967–0.970	0.926	0.850	0.954
LightGBM	Group A	0.890	0.887–0.893	0.829	0.563	0.924
	Group B	0.965	0.963–0.967	0.917	0.783	0.966