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

Table 4

Description of the optimization problems in this study.

Parameter	For XGB	For CAT
Decision variables	learning_rate	learning_rate
	max_depth	depth
	subsample	bagging_temperature
	colsample_bytree	num_leaves
	reg_alpha
	max_leaves
	gamma
	min_child_weight
Objective	Minimize(test-RMSE)	Minimize(test-RMSE)
Bounds of decision variables	0.001 ≤ learning_rate ≤ 0.3	0.001 ≤ learning_rate ≤ 0.5
	5 ≤ max_depth ≤ 30	5 ≤ depth ≤ 8
	0.5 ≤ subsample ≤ 1	3 ≤ bagging_temperature ≤ 10
	0.3 ≤ colsample_bytree ≤ 1	30 ≤ num_leaves ≤ 150
	0.005 ≤ reg_alpha ≤ 0.02
	0 ≤ max_leaves ≤ 0.02
	0 ≤ gamma ≤ 1
	1 ≤ min_child_weight ≤ 10