Table 2.

Comparison between the 1D-SE-ResNet and conventional models in terms of average Accuracy (Acc), Precision (Pre), Sensitivity (Sen), and Specificity (Spe). The standardization and smoothing is denoted as std and sm, respectively.

Model	Preprocessing	Acc of Test Set (%)								Acc (%)	Pre (%)	Sen (%)	Spe (%)
		D1	D2	D3	D4	D5	D6	D7	D8
SVM	/	92.86	96.67	79.31	78.57	100.0	92.59	90.00	93.33	90.42	94.86	86.03	95.00
	std	92.86	96.67	79.31	82.14	96.30	92.59	93.33	93.33	90.82	96.08	86.03	95.63
	sm	92.86	96.67	79.31	82.14	92.59	92.59	86.67	93.33	89.52	94.23	86.03	92.92
	PCA	92.86	96.67	79.31	75.00	100.0	92.59	90.00	93.33	89.97	93.75	86.03	94.17
RF	/	85.71	90.00	75.86	71.43	81.48	92.59	93.33	93.33	85.47	88.47	83.40	87.50
	std	85.71	90.00	75.86	71.43	81.48	92.59	93.33	93.33	85.47	88.47	83.40	87.50
	sm	85.71	93.33	75.86	75.00	85.19	96.30	100.0	93.33	88.09	92.03	85.06	90.83
	PCA	89.29	80.00	75.86	75.00	85.19	92.59	83.33	90.00	83.91	86.85	82.82	85.00
PLS-DA	/	96.43	53.33	79.31	67.86	66.67	100.0	83.33	86.67	79.20	81.11	86.03	71.46
	std	96.43	53.33	79.31	67.86	66.67	100.0	83.33	86.67	79.20	81.11	86.03	71.46
	sm	96.43	80.00	86.21	71.43	96.30	100.0	86.67	96.67	89.21	89.35	89.36	88.96
1D-SE-ResNet	/	89.29	96.67	100.0	78.57	85.19	100.0	100.0	100.0	93.72	96.06	90.77	96.25