Table 9.

Classification results for proposed method using different pre-trained CNNs and dDTF images on DEAP database in all frequency bands using LOSO Cross-Validation method

Net	Bands	Accuracy (%)	Precision (%)	Recall (%)	F1- score (%)
AlexNet	Delta	79.49 ± 3.42	79.35 ± 3.35	79.55 ± 3.72	79.32 ± 3.35
	Theta	79.36 ± 3.36	79.13 ± 3.41	79.42 ± 3.29	79.25 ± 3.26
	Alpha	80.62 ± 3.34	80.48 ± 3.30	80.72 ± 3.35	80.49 ± 3.32
	Beta	80.21 ± 3.35	80.10 ± 3.34	80.34 ± 3.42	80.15 ± 3.42
	Gamma	80.36 ± 3.30	80.24 ± 3.35	80.39 ± 3.45	80.30 ± 3.29
VGG-19	Delta	80.12 ± 3.34	80.11 ± 3.32	80.25 ± 3.35	80.08 ± 3.29
	Theta	80.46 ± 3.40	80.38 ± 3.35	80.54 ± 3.32	80.30 ± 3.35
	Alpha	81.72 ± 3.35	81.57 ± 3.26	81.76 ± 3.25	81.65 ± 3.36
	Beta	81.38 ± 3.22	81.32 ± 3.35	81.42 ± 3.27	81.32 ± 3.45
	Gamma	81.45 ± 3.36	81.34 ± 3.42	81.51 ± 3.38	81.40 ± 3.36
ResNet-50	Delta	94.02 ± 3.25	94.15 ± 3.34	94.26 ± 3.32	94.10 ± 3.35
	Theta	94.35 ± 3.27	94.25 ± 3.30	94.44 ± 3.35	94.32 ± 3.40
	Alpha	96.12 ± 3.25	96.15 ± 3.24	96.25 ± 3.30	96.15 ± 3.25
	Beta	95.42 ± 3.50	95.40 ± 3.36	95.57 ± 3.24	95.32 ± 3.22
	Gamma	95.56 ± 3.20	95.74 ± 3.29	95.53 ± 3.30	95.67 ± 3.29
Inception-v3	Delta	92.36 ± 3.55	92.23 ± 3.60	92.25 ± 3.64	92.20 ± 3.52
	Theta	92.49 ± 3.54	92.35 ± 3.62	92.35 ± 3.62	92.31 ± 3.56
	Alpha	93.80 ± 3.56	93.62 ± 3.65	93.72 ± 3.60	93.72 ± 3.52
	Beta	93.30 ± 3.52	93.13 ± 3.62	93.35 ± 3.62	93.25 ± 3.53
	Gamma	93.45 ± 3.54	93.30 ± 3.56	93.50 ± 3.62	93.35 ± 3.55

Highest results are showed bold