Table 4 Ranking of ML models based on statistical performance metrics using the hold-out validation method.
R2 | Score | MAPE | Score | RMSE | Score | VAF [%] | Score | a20-index | Score | Ranking score | |
|---|---|---|---|---|---|---|---|---|---|---|---|
DTR | 0.3401 | 1 | 0.20 | 1 | 12.33 | 1 | 60.1 | 1 | 0.64 | 1 | 5 |
SVR | 0.9647 | 12 | 0.04 | 8 | 2.85 | 12 | 98.2 | 12 | 0.96 | 7 | 51 |
NuSVR | 0.9418 | 11 | 0.05 | 7 | 3.66 | 10 | 97.1 | 10 | 0.96 | 7 | 45 |
GPR | 0.9400 | 10 | 0.05 | 7 | 3.50 | 11 | 97.3 | 11 | 0.97 | 8 | 47 |
XGBoost | 0.7391 | 2 | 0.11 | 3 | 7.75 | 2 | 86.4 | 2 | 0.82 | 2 | 11 |
RF | 0.7659 | 4 | 0.12 | 2 | 7.34 | 4 | 87.6 | 4 | 0.83 | 3 | 17 |
ETR | 0.8009 | 5 | 0.10 | 4 | 6.77 | 5 | 89.8 | 5 | 0.87 | 4 | 23 |
GBR | 0.7595 | 3 | 0.11 | 3 | 7.44 | 3 | 87.2 | 3 | 0.83 | 3 | 15 |
HGBR | 0.8060 | 6 | 0.11 | 3 | 6.69 | 6 | 89.9 | 6 | 0.87 | 4 | 25 |
ANN | 0.8947 | 8 | 0.07 | 5 | 4.92 | 8 | 94.6 | 8 | 0.95 | 6 | 35 |
VR | 0.8158 | 7 | 0.10 | 4 | 6.51 | 7 | 90.3 | 7 | 0.88 | 5 | 30 |
MLPR | 0.9126 | 9 | 0.06 | 6 | 4.49 | 9 | 95.6 | 9 | 0.97 | 8 | 41 |
