Table 5.
VIs predicts yield results.
| VIs | Algorithm | R2 | RMSE | MAE | MBE | Equation |
|---|---|---|---|---|---|---|
| EXG | BP | 0.59 | 11.02 | 14.61 | -0.18 | y=0.40x+38.51 |
| SVM | 0.61 | 15.43 | 13.32 | 1.11 | y=0.67x+23.80 | |
| RF | 0.58 | 10.78 | 11.00 | -0.02 | y=0.54x+29.02 | |
| GRAY | BP | 0.65 | 12.27 | 12.41 | -0.32 | y=0.59x+25.86 |
| SVM | 0.66 | 14.65 | 9.65 | 2.27 | y=0.77x+17.89 | |
| RF | 0.71 | 10.94 | 8.25 | 0.21 | y=0.69x+19.57 | |
| BRVI | BP | 0.61 | 11.82 | 11.06 | 0.07 | y=0.61x+25.62 |
| SVM | 0.58 | 11.07 | 9.93 | -3.15 | y=0.52x+27.34 | |
| RF | 0.68 | 11.37 | 8.74 | 0.28 | y=0.69x+20.50 | |
| NDRE | BP | 0.41 | 13.17 | 11.83 | 0.48 | y=0.47x+34.39 |
| SVM | 0.55 | 13.27 | 13.37 | -0.80 | y=0.45x+33.99 | |
| RF | 0.54 | 11.45 | 11.52 | 0.07 | y=0.47x+33.85 | |
| Combine | BP | 0.87 | 10.67 | 7.25 | -2.05 | y=0.83x+8.75 |
| SVM | 0.86 | 10.75 | 5.93 | -1.05 | y=0.83x+9.85 | |
| RF | 0.87 | 6.78 | 5.63 | -0.03 | y=0.8x+12.46 |
EXG stand for excess green index; GLI stand for green leaf index; GRAY stand for gray level index; BRVI stand for simple blue ratio index; NDRE stand for normalized difference red - edge index. BP stand for back propagation neural network; SVM stand for support vector machine; RF stand for random forest.