Table 5. Selected hyperparameters for the high accuracy classifier.

The first column contains the best hyperparameters for the “Top 13” features, as described in the previous section. The second column contains the best hyperparameters for all features generated.

Hyperparameter	Selected Values
	“Top 13”	All features
learning_rate	0.25	0.709
max_depth	4	3
min_child_weight	9.5	2.791
gamma	2.1	2.106
colsample_bytree	0.9	0.619
num_boost_round	450	336
reg_lambda	5.36	4.515
scale_pos_weight	1.0	1.437
subsample	1.0	0.504