Table 1 Hyperparameter grids explored during grid search with 5-fold cross-validation. Optimal parameters were selected based on the highest R² score on the validation folds.

From: A hybrid framework of feature selection and interpretability for dissolved oxygen prediction in drinking water treatment plants

Model

Parameter

Searched Values

Random Forest

n_estimators

[50, 100, 200]

 

max_depth

[None, 10, 20]

 

min_samples_split

[2, 5, 10]​

 

max_features

[‘sqrt’, ‘log2’]

XGBoost

learning_rate

[0.01, 0.1, 0.3]

 

max_depth

[3, 6, 10]

 

n_estimators

[50, 100, 200]​

 

subsample

[0.6, 0.8, 1.0]

Back to article page