Skip to main content
. 2018 Jun;227:18–29. doi: 10.1016/j.jfoodeng.2018.01.009

Table 3.

Performance of the PLS regression model for total lipid content in green coffee beans expressed on an “as is” or dry matter basis.

Pre-processing LV R2 Cal RMSEC Slope R2 Val RMSECV Slope RPD
As-is
Log (1/R) 6 0.850 1.019 0.850 0.837 1.062 0.843 2.49
First derivative 4 0.866 0.961 0.866 0.860 0.985 0.857 2.68
Second derivative 6 0.893 0.861 0.893 0.883 0.904 0.885 2.92
SNV 5 0.859 0.990 0.859 0.849 1.028 0.857 2.57
Baseline
6
0.852
1.010
0.852
0.841
1.050
0.847
2.51
Dry matter basis Log (1/R) 6 0.864 1.087 0.864 0.855 1.125 0.862 2.76
First derivative 4 0.871 1.062 0.871 0.866 1.091 0.864 2.84
Second derivative 6 0.900 0.935 0.900 0.890 0.985 0.890 3.15
SNV 5 0.864 1.081 0.864 0.855 1.117 0.860 2.78
Baseline 6 0.857 1.116 0.857 0.845 1.169 0.853 2.65

Spectral range used: 980-2480 nm. LV: latent variable. SNV: standard normal variate spectral treatment. Error is indicated as %. Sample size (n): 350.