Table 3.

represents the 3-Class Fold wise analysis.

FOLD	METHOD	Precision	Recall	F-score	Accuracy
Fold 1	MobileNetV2	0.9525	0.9502	0.9512	0.9484
	VGG16	0.9527	0.9522	0.9523	0.9509
	Fusion	0.9677	0.9663	0.9668	0.9656
Fold 2	MobileNetV2	0.9476	0.9501	0.9487	0.9476
	VGG16	0.9579	0.957	0.9572	0.9558
	Fusion	0.9658	0.9676	0.9666	0.9656
Fold 3	MobileNetV2	0.9461	0.9454	0.9458	0.9443
	VGG16	0.9445	0.9412	0.9427	0.9410
	Fusion	0.9611	0.959	0.96	0.9590
Fold 4	MobileNetV2	0.9379	0.9388	0.9383	0.9369
	VGG16	0.9537	0.9534	0.9535	0.9525
	Fusion	0.969	0.9699	0.9694	0.9689
AVERAGE	MobileNetV2	0.9460	0.9461	0.9460	0.9443
	VGG16	0.9522	0.95095	0.9514	0.9500
	Fusion	0.9659	0.9657	0.9657	0.9648