Research Article
Feasibility Study on the Use of a Portable NIR Spectrometer and Multivariate Data Analysis to Discriminate and Quantify Adulteration in Fertilizer
Table 4
Optimum performance of different quantification models for fertilizer A.
| Model | R2 | RMSECV | Bias | R2 | RMSEP | Bias | RPD |
| PLS | 0.9351 | 0.1297 | 0.0094 | 0.9450 | 0.1201 | 0.0133 | 3.0767 | iPLS | 0.9346 | 0.1302 | −0.0003 | 0.9433 | 0.1227 | −0.0045 | 3.0128 | Bi-PLS | 0.9424 | 0.1224 | −0.0005 | 0.9299 | 0.1362 | −0.0134 | 2.7141 | Si-PLS | 0.9496 | 0.1147 | 0.0004 | 0.9529 | 0.1122 | −0.0076 | 3.2947 | GAPLS | 0.9409 | 0.1240 | −0.0011 | 0.9493 | 0.1157 | 0.0010 | 3.1950 |
| | Variables | Selected wavelength range (nm) | PLS | 331 | 740–1070 | iPLS | 16 | 927–942 | Bi-PLS | 50 | 825–841, 927–942, 959–975 | Si-PLS | 58 | 820–839, 938–956, 995–1013 | GAPLS | 331 | 740–1070 |
|
|