Table 3.

Performance of feature-based models. Three feature-based models were tested using 6 evaluation metrics.

Model	AUC	Accuracy	MCC	Precision	Recall	F1
Support vector machines (svm)	0.74	68.66%	0.37	0.69	0.69	0.69
Random forest	0.73	67.24%	0.34	0.66	0.68	0.67
Logistic regression	0.73	67.35%	0.39	0.70	0.69	0.69