Table 3. Hyperparameters optimization search space for the tested ML models.

ML algorithm	Hyperparameters	Search space [lower, upper values]; step
J48 (grid search)	Maximum depth	[2, 10]; 1
	Minimum sample split	[5, 20]; 2
XGBoost (genetic algorithm)	Maximum tree depth	[3, 15]; 2
	Learning rate (eta)	[0.10, 0.30]; 0.05
	Alpha	[0, 5]; 1
	Lambda	[0, 5]; 1
	Gamma	[0, 5]; 1
	Minimum child weight	[1,10]; 2
	Subsampling rate	[0.5, 1.0]; 0.1
kNN (grid search)	Number of neighbours (k)	[3, 20]; 1