Table 2 Specific results and parameters of the data-driven model.
Model | SD and CV(%) | Fold status | Folds | Avg loss | Key parameters |
|---|---|---|---|---|---|
LASSO | 0.0032, 0.73% | Best | 6 | 0.441 | Alpha: 0.00214 |
Worst | 5 | 0.4471 | Alpha: 0.00351 | ||
XGBoost | 0.002, 0.48% | Best | 3 | 0.4531 | LR: 0.0611, Depth: 4, Trees: 270 |
Worst | 7 | 0.4574 | LR: 0.1077, Depth: 3, Trees: 250 | ||
RF | 0.0102, 2.67% | Best | 4 | 0.414 | Depth: 20, Split: 0.1013, Trees: 160 |
Worst | 5 | 0.4389 | Depth: 15, Split: 0.1004, Trees: 300 | ||
NN | 0.0275, 4.68% | Best | 4 | 0.4479 | Act: logistic, Alpha: 0.090, Layers: 100, Iter: 700 |
Worst | 3 | 0.5169 | Act: relu, Alpha: 0.00013, Layers: 50, Iter: 900 | ||
SVM | 0.0048, 0.98% | Best | 6 | 0.4423 | C: 63.62, Epsilon: 0.5525 |
Worst | 2 | 0.4544 | C: 87.66, Epsilon: 0.3909 |
