Peaberry coffee discrimination using UV-visible spectroscopy combined with SIMCA and PLS-DA

Figures & data

Figure 1. Original spectra of peaberry and normal coffee samples in ultraviolet-visible region (190–700 nm).

Figure 2. Principal components analysis of spectra from the peaberry and normal coffee samples (projection onto the first two principal components).

Figure 3. Cooman plot for the classification of peaberry and normal coffee samples (vertical and horizontal red lines show 95% confidence interval).

Table 1. Confusion matrix with accuracy, sensitivity, and specificity of the classification of coffee samples by SIMCA using the prediction sample set.

	Class peaberry coffee (assigned by the SIMCA model peaberry)	Class normal coffee (assigned by the SIMCA model normal)
Class peaberry coffee (actual)	16	0
Class normal coffee (actual)	0	16
Accuracy (%)	100
Sensitivity (%)	100
Specificity (%)	100

Figure 4. Estimated class values by the PLS-DA model versus the calibration and prediction sample plots for discrimination between peaberry and normal coffees.

Figure 5. X-Loading weights versus wavelength of the top two PCs (PC1 and PC2) of coffee samples (upper: peaberry coffee; lower: normal coffee).

Figure 6. Plots of loadings of latent variables versus wavelength for the PLS-DA model.