Table 4.

GBM hyperparameters

Hyperparameter	Value/Range
n_estimators	[100, 1000] (step = 100)
learning_rate	[1e−8, 1.0] (log scale)
max_depth	[2, 12]
gamma	[1e−8, 1.0] (log scale)
min_child_weight	[1e−8, 1e2] (log scale)
subsample	[0.4, 1.0]
reg_lambda	[1e−6, 10] (log scale)

Best parameters differ for each benchmarking dataset and split