Table 2 Soybean yield estimation performance of the GCBA and comparing models.
From: GOA-optimized deep learning for soybean yield estimation using multi-source remote sensing data
Year | Model | Training dataset | Testing dataset | ||||||
|---|---|---|---|---|---|---|---|---|---|
R2 | RMSE (bushels/acre) | MAE (bushels/acre) | MAPE (%) | R2 | RMSE (bushels/acre) | MAE (bushels/acre) | MAPE (%) | ||
2019 | SVR | 0.6672 | 5.3363 | 3.9256 | 8.79 | 0.4675 | 5.6489 | 4.1651 | 8.48 |
RFR | 0.8844 | 3.1936 | 2.2666 | 5.26 | 0.5320 | 5.2956 | 4.0861 | 8.71 | |
CNN | 0.8906 | 3.0602 | 2.0637 | 4.53 | 0.5840 | 4.9927 | 3.7847 | 7.89 | |
GRU | 0.9035 | 2.8733 | 1.9392 | 4.25 | 0.6391 | 4.6507 | 3.5475 | 7.31 | |
CNN-GRU | 0.9114 | 2.7541 | 1.8465 | 4.04 | 0.6633 | 4.4919 | 3.4383 | 7.27 | |
GCBA | 0.9203 | 2.6121 | 1.7547 | 3.83 | 0.6873 | 4.3288 | 3.2712 | 6.88 | |
2020 | SVR | 0.6743 | 5.2793 | 3.8565 | 8.63 | 0.4082 | 6.3444 | 4.1689 | 8.46 |
RFR | 0.8295 | 3.8202 | 2.6299 | 5.77 | 0.5364 | 5.5906 | 3.7262 | 7.65 | |
CNN | 0.8143 | 3.9865 | 2.7777 | 6.11 | 0.6152 | 5.0943 | 3.1645 | 6.32 | |
GRU | 0.8616 | 3.4415 | 2.3305 | 5.10 | 0.6567 | 4.8264 | 3.0281 | 6.08 | |
CNN-GRU | 0.9032 | 2.8788 | 1.9162 | 4.16 | 0.6671 | 4.7926 | 2.9962 | 5.93 | |
GCBA | 0.9255 | 2.5251 | 1.6805 | 3.62 | 0.7057 | 4.4612 | 2.8684 | 5.80 | |
