Research Article
Feasibility Study on the Use of a Portable NIR Spectrometer and Multivariate Data Analysis to Discriminate and Quantify Adulteration in Fertilizer
Table 5
Optimum performance of different quantification models for fertilizer B.
| Model | R2 | RMSECV | Bias | R2 | RMSEP | Bias | RPD |
| PLS | 0.9523 | 0.0801 | 0.0054 | 0.9529 | 0.0898 | 0.0092 | 4.3444 | iPLS | 0.9613 | 0.0945 | 0.0004 | 0.9648 | 0.0908 | −0.0007 | 3.8200 | Bi-PLS | 0.9664 | 0.0881 | −0.0002 | 0.9676 | 0.0879 | −0.0050 | 3.9460 | Si-PLS | 0.9767 | 0.0736 | 0.0007 | 0.9799 | 0.0692 | −0.0012 | 5.0124 | GAPLS | 0.9736 | 0.0782 | 0.0004 | 0.9766 | 0.0742 | 0.0028 | 4.6746 |
| | Variables | Selected wavelength range (nm) | PLS | 331 | 740–1070 | iPLS | 17 | 757–773 | Bi-PLS | 47 | 757–773, 959–974, 1009–1022 | Si-PLS | 83 | 740–767, 824–851, 990–1016 | GAPLS | 331 | 740–1070 |
|
|