Table 3.

The comparison of the prediction model developed on the original and balance dataset

Prediction model		Dataset	Accuracy	Precision	Recall	f1-score	AUC
Random forest model	RF-Model-O	Internal validation	0.83	0.84	0.82	0.83	0.87
	RF-Model-B		0.80	0.80	0.80	0.80	0.87
	RF-Model-O	External validation	0.68	0.66	0.68	0.67	0.69
	RF-Model-B		0.71	0.69	0.71	0.69	0.69
Support vector	SV-Model-O	Internal validation	0.78	0.80	0.78	0.78	0.82
	SV-Model-B		0.75	0.76	0.75	0.75	0.83
	SV-Model-O	External validation	0.57	0.62	0.57	0.58	0.61
	SV-Model-B		0.61	0.62	0.61	0.61	0.61
Gradient boost	GB-Model-O	Internal validation	0.80	0.81	0.80	0.80	0.81
	GB-Model-B		0.80	0.82	0.80	0.80	0.86
	GB-Model-O	External validation	0.68	0.66	0.68	0.67	0.72
	GB-Model-B		0.65	0.63	0.65	0.64	.65

O stands for original data set, B stands for balanced dataset