Rapid and Nondestructive Identification of Origin and Index Component Contents of Tiegun Yam Based on Hyperspectral Imaging and Chemometric Method

Table 3

Pairwise combination classification accuracy of the preprocessing method and classification model of Tiegun yam powder samples from different origins.

Models	Preprocessing	Accuracy (%)
		Training set	Prediction set
PLS-DA	Raw data	99.57	96.00
	MSC	100.00	98.40
	D1	100.00	96.00
	D2	97.82	95.42
	SG	98.72	95.20
	SNV	99.15	96.80
SVM	Raw data	98.81	61.11
	MSC	97.22	61.11
	D1	98.81	25.93
	D2	34.52	27.78
	SG	98.41	59.26
	SNV	100.00	24.07
RF	Raw data	97.14	68.89
	MSC	99.05	76.67
	D1	99.05	52.22
	D2	100.00	83.33
	SG	98.09	77.78
	SNV	97.14	68.89