Table 5.

Overview of the RF hyperparameters and their values included in optimization.

Hyperparameter	Description	Evaluated values
Mtry	Number of randomly drawn candidate variables in each split for growing a tree	$\sqrt{V}$, 25 %, 33.3 % and 40 % of V
Node size (minimum.node.size)	Minimum number of observations in a terminal node	1, 3 and 5
Replace	Sampling approach	TRUE (sample with replacement) and FALSE (sample without replacement)
Sample.fraction	Fraction of observations in the calibration dataset to sample in each tree	0.50, 0.63 and 0.80

V: number of predictor variables.