Table 6. Classification results on the fused Alexnet, VGG19, Inceptionv3 DCNN features after entropy based selection. Best values are shown in bold.

Classifier	Performance measures
	Sensitivity (%)	Specificity (%)	FNR (%)	Accuracy (%)	F-score
Cubic SVM	98.2	97.5	2.3	97.6	0.979
Linear SVM	95.16	91.7	4.8	93.5	0.935
Quadratic SVM	96.50	95.5	3.4	96.0	0.960
Fine KNN	93.5	94.0	6.4	93.8	0.937
Medium KNN	96.7	90.68	3.1	93.8	0.939
Cubic KNN	96.41	89.6	3.5	93.1	0.932
Weighted KNN	96.31	89.2	3.3	93.1	0.930