TABLE 7.

Secondary outcome classification (SG vs. TG) based on different machine learning models.

Classifier	Model	AUC	Accuracy	Precision	Recall	MCC	Specificity
KNN	CV-2	0.538	0.476	0.469	0.476	−0.031	0.493
AB		0.534	0.540	0.533	0.540	0.095	0.554
NB		0.512	0.270	0.514	0.270	0.022	0.761
RF		0.475	0.444	0.434	0.444	−0.095	0.462
KNN	CV-5	0.563	0.508	0.502	0.508	0.032	0.523
AB		0.421	0.429	0.421	0.429	−0.125	0.446
NB		0.490	0.095	0.410	0.095	−0.045	0.880
RF		0.604	0.571	0.562	0.571	0.157	0.585
KNN	CV-10	0.591	0.571	0.574	0.571	0.166	0.585
AB		0.526	0.540	0.531	0.540	0.094	0.554
NB		0.487	0.048	0.246	0.048	−0.066	0.924
RF		0.550	0.508	0.503	0.508	0.032	0.523
Mean		0.524	0.416	0.474	0.416	0.019	0.607
±SD		0.049	0.173	0.085	0.173	0.091	0.152

Bold values mean highest value.