Table 1.

The performance of models before and after adjusting the parameters.

	before			after
	Train AUC	Test AUC	Validation AUC	Train AUC	Test AUC	Validation AUC
F-Lasso	86.8%	84.5%	49.9%	86.9%	84.8%	50.6%
T-SVM	88.1%	83.5%	52.1%	88.4%	84.5%	54.3%
RF-RF	89.4%	89.7%	52.6%	90.3%	90.6%	49.4%
RF-GBRT	89.5%	89.9%	49.9%	90.8%	91.1%	55.5%
Chi2-GBRT	86.5%	86.8%	62.7%	87.3%	87.5%	70.7%
Log-GBRT	85.2%	85.4%	65.1%	85.9%	86.2%	70.8%