Table 15.

The result of ResNet18 for classification using tenfold cross-validation

	NE. (%)	EO. (%)	BA. (%)	LY. (%)	MO. (%)
Fold-0	96	99	53	61	76
Fold-1	83	94	97	97	89
Fold-2	54	21	86	73	91
Fold-3	89	92	70	68	84
Fold-4	74	92	100	85	82
Fold-5	86	82	56	74	69
Fold-6	84	93	100	88	62
Fold-7	70	91	90	99	93
Fold-8	61	81	61	73	42
Fold-9	97	86	100	83	71
Avr. Acc.	79	83	81	80	75

The average accuracy for five classes is 79%