Table 3.

Classification results of 5 DCNN classifiers trained on 256 × 256 patch size for each cross validation in 3 different testing folds

	Classification results for 4 classes
	Index (avg. on 4 classes)	Fold #1	Fold #2	Fold #3	Average
VGG19	Recall	0.61	0.83	0.51	0.65
	Precision	0.61	0.84	0.51	0.66
	F1-Score	0.61	0.83	0.50	0.65
	MCC	0.17	0.56	− 0.05	0.23
MobileNetV2	Recall	0.86	0.85	0.87	0.86
	Precision	0.85	0.85	0.87	0.86
	F1-Score	0.85	0.85	0.87	0.86
	MCC	0.67	0.66	0.68	0.67
ResNet50	Recall	0.84	0.86	0.60	0.77
	Precision	0.85	0.86	0.64	0.78
	F1-Score	0.84	0.86	0.60	0.58
	MCC	0.65	0.68	0.10	0.36
InceptionV3	Recall	0.87	0.87	0.87	0.87
	Precision	0.86	0.87	0.87	0.87
	F1-Score	0.87	0.87	0.87	0.87
	MCC	0.69	0.69	0.70	0.70
GoogLeNet	Recall	0.99	0.98	0.98	0.98
	Precision	0.98	0.98	0.99	0.98
	F1-Score	0.98	0.99	0.99	0.98
	MCC	0.97	0.96	0.97	0.97