Table 6 Statistical evaluation of the developed models.
From: Modeling liquid rate through wellhead chokes using machine learning techniques
| Â | Adaboost-SVR | MARS | MLP-LM | MLP-BR | MLP-SCG | RBF |
|---|---|---|---|---|---|---|
Training set | ||||||
AAPRE % | 5.3 | 6.58 | 8.7 | 9.24 | 11.51 | 8.11 |
APRE % | − 1.76 | − 1.19 | − 2.66 | − 2.42 | − 2.04 | − 1.84 |
RMSE | 661.56 | 469.19 | 682.99 | 747.87 | 917.04 | 672.42 |
SD | 0.09 | 0.149 | 0.31 | 0.31 | 0.3 | 0.22 |
R2 | 0.9772 | 0.9889 | 0.9757 | 0.9707 | 0.9525 | 0.9761 |
Test set | ||||||
AAPRE % | 4.57 | 12.14 | 9.19 | 7.92 | 10.9 | 8.45 |
APRE % | − 0.47 | − 4.27 | − 2.08 | − 0.88 | − 3.13 | 1.76 |
RMSE | 564.86 | 921.01 | 913.49 | 710.84 | 976.23 | 959.29 |
SD | 0.1 | 0.31 | 0.14 | 0.13 | 0.19 | 0.14 |
R2 | 0.9827 | 0.9465 | 0.9571 | 0.9744 | 0.9477 | 0.9559 |
Total | ||||||
AAPRE % | 5.15 | 7.69 | 8.9 | 8.74 | 11.44 | 8.18 |
APRE % | − 1.5 | − 1.81 | − 2.3 | − 1.99 | − 1.97 | − 1.12 |
RMSE | 643.38 | 588.02 | 726.34 | 733.78 | 958.04 | 738.76 |
SD | 0.086 | 0.19 | 0.29 | 0.28 | 0.28 | 0.2 |
R2 | 0.9784 | 0.9819 | 0.9726 | 0.9719 | 0.9522 | 0.9716 |
