Table 1.

Hyper-parameters fine-tuned using Optuna for each model.

Algorithm	Fine-Tuned Hyper-Parameters
AdaBoost	Learning rate; loss; number of estimators
SVR	c; γ; epsilon
Lasso	α; maximum number of iterations
Ridge	α; maximum number of iterations
PLSR	Number of components
RF	Maximum number of features; maximum depth; minimum samples to split an internal node; minimum number of samples of a leaf node after the split; number of estimators
XGBoost	Learning rate; γ; minimum child weight; column sample by tree; subsample; maximum depth