Solving the class imbalance problem using ensemble algorithm: application of screening for aortic dissection

Table 5 Sensitivity (Se) and specificity (Sp) of SVM models with different weights on positive and negative samples

	SVM (1,1)		SVM (1.3,1)		SVM (1.6,1)		SVM (2,1)
	Se	Sp	Se	Sp	Se	Sp	Se	Sp
1st	0.772	0.792	0.825	0.746	0.842	0.697	0.868	0.653
2nd	0.746	0.807	0.754	0.751	0.754	0.691	0.789	0.669
3rd	0.781	0.768	0.816	0.727	0.860	0.675	0.868	0.644
4th	0.746	0.790	0.781	0.751	0.807	0.696	0.816	0.666
5th	0.772	0.805	0.781	0.756	0.798	0.684	0.851	0.646
6th	0.728	0.795	0.763	0.741	0.781	0.687	0.833	0.648
7th	0.771	0.779	0.847	0.727	0.873	0.679	0.898	0.642
Average	0.759	0.791	0.795	0.734	0.816	0.687	0.846	0.653

ISSN: 1472-6947