BMC Bioinformatics

Table 2 Performance of models based on both positive dataset and negative 2 dataset using SVM as classifier

From: Incorporating support vector machine with sequential minimal optimization to identify anticancer peptides

Features	Training				Testing
Features	Weight	Sensitivity	Specificity	Accuracy	Sensitivity	Specificity	Accuracy
AAC	0.6	0.873	0.952	0.913	0.860	0.942	0.928
N5C5	1.9	0.942	0.909	0.925	0.960	0.878	0.891
k-space = 0	0.4	0.799	0.957	0.878	0.750	0.952	0.918
k-space = 1	0.4	0.810	0.948	0.879	0.700	0.934	0.895
k-space = 2	0.5	0.840	0.957	0.898	0.720	0.958	0.918
AAC + k-space = 0	0.9	0.868	0.920	0.894	0.860	0.914	0.905
AAC + N5C5	0.4	0.892	0.972	0.932	0.940	0.952	0.950
N5C5 + k-space = 0	0.5	0.857	0.957	0.907	0.890	0.930	0.923
AAC + N5C5 + k-space 0	1.0	0.909	0.909	0.909	0.930	0.900	0.905
PSSM	0.7	0.909	0.911	0.913	0.910	0.838	0.850

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com