Applying Bayesian Optimization for Machine Learning Models in Predicting the Surface Roughness in Single-Point Diamond Turning Polycarbonate

Table 6

Optimized hyperparameters for XGBoost and CAT.

Algorithm	Library	Hyperparameters tuned	Domain	Optimized value
XGB	XGBoost	n_estimators	100	100
		learning_rate	(0.001, 0.3)	0.1220
		max_depth	(5, 30)	9
		subsample	(0.5, 1)	0.7433
		colsample_bytree	(0.3, 1)	0.7182
		reg_alpha	(0.005, 0.02)	0.0090
		max_leaves	(0, 0.02)	0
		gamma	(0, 1)	0.3778
		min_child_weight	(1, 10)	3.9175
CAT	catboost	iterations	(1000, 2000)	1000
		depth	(5, 8)	6.1114
		learning_rate	(0.001, 0.5)	0.4669
		bagging_temperature	(3, 10)	6.8176
		num_leaves	(30, 150)	138