Table 4.

Classification accuracy on the DB5 Dataset. (We use the same samples to test these methods).

Model	TD+LDA	TD+SVM	Enhanced TD+LDA	Enhanced TD+SVM	NinaPro+LDA	NinaPro+SVM	SampEn+LDA	SampEn+SVM	CNN_LSTM	LCNN	CWT+EMGNet
EA(%)	61.61	56.40	64.36	58.32	61.66	57.12	64.53	55.28	61.21	66.38	69.62
EB(%)	53.80	50.70	57.80	51.17	54.05	51.18	58.20	55.34	60.46	62.51	67.42
EC(%)	43.48	46.49	48.26	47.18	44.61	43.46	47.89	38.12	52.99	55.87	61.63
$\frac{\mathrm{EA}-\mathrm{EC}}{\mathrm{EA}}$	29.4%	17.5%	25.0%	19.1%	29.5%	23.9%	25.8%	31.0%	13.4%	15.8%	11.5%

¹ EA, EB and EC represent the accuracy of exercise A, exercise B and exercise C respectively.