Table 3 Results of Bayesian Optimization with fivefold cross-validation, for XGB-FS and XGB-FE models.

From: Application of XGBoost model for in-situ water saturation determination in Canadian oil-sands by LF-NMR and density data

XGBoost hyperparameters

Search range

XGB-FS optimal

XGB-FE optimal

n_estimators

[50–1000]

[650]

[300]

learning_rate

[0.004–0.1]

[0.008]

[0.053]

subsample

[0.7–1.0]

[0.7]

[0.6]

max_depth

[6–12]

[8]

[7]

objective

[‘squared_error’, ‘pseudo_huber’]

[‘pseudo_huber’]

[‘squared_error’]

grow_policy

[‘depthwise’, ‘lossguide’]

[‘lossguide’]

[‘lossguide’]

booster

[‘gbtree’, ‘dart’]

[‘gbtree’]

[‘gbtree’]

Back to article page