Table 3. Comparison of models using different spectral preprocessing methods to evaluate the PLSR model

	Pre-processing method¹⁾
	SNV		MSC		SG-1		SG-2
	R_P^{2 2)}	RMSEP³⁾	R_P²	RMSEP	R_P²	RMSEP	R_P²	RMSEP
Capsaicinoids	0.7560	113.2623	0.7809	109.504	0.7788	117.6174	0.3275	210.8753
Water contents	0.6634	1.0433	0.6811	0.9995	0.6979	0.9982	0.5475	1.2035
ASTA color value	0.8892	5.4664	0.8984	5.2210	0.8413	6.5242	0.6025	10.6200
Total sugar	0.5799	1.9382	0.5677	1.9662	0.5244	2.0767	0.2208	2.6644

SNV, standard normal variate; MSC, multiple scatter correction; SG-1, Savitzky-Golay 1st derivative; SG-2, Savitzky-Golay 2nd derivative.
R², coefficient of the prediction.
RMSEP, root mean square error of prediction.