Regression methods | AGB | PB |
---|
r2 | RMSEP (kg/m2) | RRMSE (%) | r2 | RMSEP (kg/m2) | RRMSE (%) |
---|
RF | 0.90 | 0.21 | 13.56 | 0.64 | 0.11 | 14.14 |
ELM | 0.87 | 0.22 | 15.42 | 0.64 | 0.11 | 14.16 |
BPNN | 0.87 | 0.23 | 15.73 | 0.59 | 0.13 | 15.85 |
LS-SVM | 0.89 | 0.21 | 14.64 | 0.64 | 0.11 | 14.14 |
- RF, ELM, BPNN and LS-SVM represent random forest, extreme learning machine, back propagation neural network and least square-support vector machine. The r2, RMSEP and RRMSE represent the coefficient of determination, the prediction of root mean square error and relative RMSE, respectively