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

Table 4

Prediction results of spectral data of Tiegun yam powder samples after different pretreatments.

Content	Models	Training set		Prediction set
		R^2	RMSEC (mg/g)	R^2	RMSEP (mg/g)	RPD
Starch	PLSR	0.9598	11.0908	0.9270	13.6931	3.22
	SVR	0.9948	3.8399	0.9636	10.7103	5.21
	RF	0.9896	8.4654	0.9677	14.4532	3.45
Polysaccharide	PLSR	0.8684	2.6837	0.8425	2.9637	2.52
	SVR	0.9996	0.1397	0.9275	2.0717	3.21
	RF	0.9818	1.6933	0.9012	3.5702	1.45
Protein	PLSR	0.8933	2.4077	0.8610	2.6676	2.61
	SVR	0.9943	0.5219	0.8939	2.6464	2.94
	RF	0.9770	1.7388	0.9292	3.5203	1.63