Skip to main content
. 2021 Feb 10;7:e386. doi: 10.7717/peerj-cs.386

Table 4. Classification results on VGG19 DCNN features. Best values are shown in bold.

Classifier Performance measures
Sensitivity (%) Specificity (%) FNR (%) Accuracy (%) F-score
Quadratic SVM 92.2 90.8 7.7 91.5 0.916
Linear SVM 90.8 87.6 9.1 89.2 0.894
Cubic SVM 93.4 91.6 6.6 92.5 0.926
Fine KNN 88.1 90.6 11.8 89.4 0.892
Medium KNN 92.2 89.2 7.7 90.7 0.908
Cubic KNN 92.4 87.7 7.5 90.1 0.903
Weighted KNN 90.8 91.3 9.1 91.1 0.910
Ensemble subspace discriminant 89.4 89.3 10.5 89.4 0.894