Table 2.

Hyperparameter search spaces for tuning the XGBoost classifier algorithm.

Hyperparameter	Description	Search space
eta (learning rate)	Learning rate to scale the contribution of each tree	[0.001, 0.01, 0.1, 0.3, 0.5]
n_estimators	Number of boosting rounds (trees) to build	[32, 64, 128, 192, 256, 384, 512]
gamma	Minimum loss reduction required to perform a split	[0, 0.25, 0.5, 1]
max_depth	Maximum depth of trees to prevent overfitting	[2, 3, 4, 6, 8, 10, 12, 16, 24]
min_child_weight	Minimum sum of instance weights needed in a child node	[0.5, 1, 3, 5, 7, 10]
subsample	Fraction of training data used for building each tree	[0.8, 0.9, 1.0]
colsample_bytree	Fraction of features randomly sampled for each tree	[0.6, 0.7, 0.8, 0.9]
lambda (reg_lambda)	L2 regularization to penalize large weights	[0.01, 0.1, 1, 5, 10, 50, 100]
alpha (reg_alpha)	L1 regularization to encourage sparsity in feature weight	[0, 0.001, 0.01, 0.1]