Table 1 RSME, R2, MAE, and Pearson’s correlation of ML-BA models
Model | Training set (80%) | Test set (20%) | ||||||
|---|---|---|---|---|---|---|---|---|
RMSE | R2 | MAE | Pearson’s correlation | RMSE | R2 | MAE | Pearson’s correlation | |
Stacking (SVM) | 5.765 | 0.438 | 4.349 | 0.661 | 5.776 | 0.435 | 4.352 | 0.659 |
Stacking (GAM) | 5.777 | 0.434 | 4.409 | 0.658 | 5.774 | 0.433 | 4.403 | 0.658 |
Stacking (MLR) | 5.788 | 0.431 | 4.418 | 0.657 | 5.786 | 0.431 | 4.414 | 0.656 |
Stacking (RF) | 2.786 | 0.900 | 2.094 | 0.949 | 5.828 | 0.422 | 4.444 | 0.650 |
XGBoost | 4.988 | 0.578 | 3.780 | 0.760 | 5.869 | 0.414 | 4.489 | 0.643 |
CatBoost | 3.674 | 0.771 | 2.739 | 0.878 | 5.893 | 0.409 | 4.494 | 0.640 |
LGBM | 4.128 | 0.711 | 3.097 | 0.843 | 5.926 | 0.403 | 4.538 | 0.634 |
GBDT | 5.513 | 0.484 | 4.239 | 0.696 | 5.951 | 0.397 | 4.579 | 0.630 |
Extra Trees | 0.000 | 1.000 | 0.000 | 1.000 | 6.319 | 0.321 | 4.889 | 0.566 |
DNN | 6.251 | 0.341 | 4.869 | 0.584 | 6.419 | 0.299 | 5.014 | 0.547 |
CNN | 5.918 | 0.409 | 4.583 | 0.640 | 6.467 | 0.289 | 5.016 | 0.537 |
GAM | 6.516 | 0.279 | 5.094 | 0.529 | 6.509 | 0.280 | 5.072 | 0.529 |
MLR | 6.692 | 0.240 | 5.238 | 0.490 | 6.691 | 0.239 | 5.224 | 0.489 |
AdaBoost | 6.986 | 0.172 | 5.499 | 0.414 | 6.994 | 0.168 | 5.501 | 0.409 |