Table 2.
Prediction results of PLSR models of Tartary buckwheat from Inner Mongolia, Sichuan and Shanxi under different preprocessing after CARS screening.
| Preprocessing methods | Inner Mongolia |
Sichuan |
Shanxi |
|||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Number of variables | Optimal factor number | Calibration set |
Prediction set |
Number of variables | Optimal factor number | Calibration set |
Prediction set |
Number of variables | Optimal factor number | Calibration set |
Prediction set |
|||||||
| R2c | RMSEC | R2p | RMSEP | R2c | RMSEC | R2p | RMSEP | R2c | RMSEC | R2p | RMSEP | |||||||
| RAW | 26 | 9 | 0.9986 | 0.0173 | 0.9984 | 0.0182 | 35 | 10 | 0.9923 | 0.0406 | 0.9879 | 0.0480 | 39 | 9 | 0.9970 | 0.0251 | 0.9975 | 0.0234 |
| SNV | 35 | 8 | 0.9988 | 0.0160 | 0.9985 | 0.0174 | 37 | 8 | 0.9919 | 0.0399 | 0.9913 | 0.0424 | 39 | 7 | 0.9967 | 0.0254 | 0.9970 | 0.0263 |
| MSC | 29 | 6 | 0.9987 | 0.0166 | 0.9987 | 0.0167 | 35 | 8 | 0.9917 | 0.0404 | 0.9924 | 0.0400 | 39 | 7 | 0.9967 | 0.0256 | 0.9976 | 0.0240 |
| MMN | 32 | 7 | 0.9987 | 0.0168 | 0.9988 | 0.0152 | 35 | 10 | 0.9916 | 0.0408 | 0.9900 | 0.0438 | 35 | 8 | 0.9966 | 0.0260 | 0.9978 | 0.0227 |
| SG | 32 | 10 | 0.9986 | 0.0174 | 0.9775 | 0.0669 | 39 | 10 | 0.9921 | 0.0411 | 0.9903 | 0.0430 | 29 | 9 | 0.9968 | 0.0257 | 0.9975 | 0.0234 |
| SG-1st | 44 | 7 | 0.9984 | 0.0186 | 0.9633 | 0.0868 | 47 | 27 | 0.9908 | 0.0427 | 0.9805 | 0.0550 | 42 | 5 | 0.9965 | 0.0263 | 0.9924 | 0.0373 |
| SNV-DT | 41 | 7 | 0.9987 | 0.0163 | 0.9984 | 0.0175 | 55 | 8 | 0.9914 | 0.0412 | 0.9902 | 0.0432 | 51 | 8 | 0.9969 | 0.0249 | 0.9976 | 0.0235 |