Table 3. Statistical results of the average of 10 iterations of the three models.

Figures in parentheses are the corresponding variance of the 10 independent results.

	Dataset	F1 score-DTPs	Recall Ratio-DTPs	Precision-NDTPs
SAE-BSVM	Training	0.349(0.179)	1(0)	1(0)
	Testing	0.234(0.013)	0.712(0.101)	0.587(0.027)
Wrapper-BSVM	Training	0.179(0.087)	0.482(0.131)	0.926(0.006)
	Testing	0.169(0.016)	0.451(0.141)	0.923(0.006)
Origin-BSVM	Training	1(0)	1(0)	1(0)
	Testing	0(0)	0(0)	0.91(0)