Prediction of Soil Available Boron Content in Visible-Near-Infrared Hyperspectral Based on Different Preprocessing Transformations and Characteristic Wavelengths Modeling
Table 9
Result of MSC/BPNN models.
Variable selection method
No. of variables
Calibration sets
Test sets
RMSE
RPD
RPIQ
RMSE
RPD
RPIQ
Full wavelengths
1306
0.764
0.415
2.066
1.129
0.816
0.379
2.349
1.054
RFA
328
0.788
0.394
2.179
1.191
0.841
0.352
2.530
1.136
SPA
24
0.733
0.441
1.944
1.063
0.736
0.453
1.964
0.882
VIP
108
0.740
0.436
1.967
1.075
0.832
0.361
2.463
1.106
Note. Bold indicates that the prediction accuracy of the model is good.