A PROC HPFOREST model is trained with VARS_TO_TRY=5 on a dataset with 40 predictor variables. What does VARS_TO_TRY=5 control?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Here's the deal — b is correct because VARS_TO_TRY= controls the random feature subspace at each node: only 5 of the 40 available predictors are randomly selected and evaluated for the optimal split at each node. This is the defining mechanism of random forests that creates decorrelated trees.
Full explanation below image
Full Explanation
B is correct because VARS_TO_TRY= controls the random feature subspace at each node: only 5 of the 40 available predictors are randomly selected and evaluated for the optimal split at each node. This is the defining mechanism of random forests that creates decorrelated trees. The typical default is sqrt(p) for classification problems. A (leaf size) is controlled by LEAFSIZE= or MINLEAFSIZE=. C describes a different concept unrelated to VARS_TO_TRY. D is wrong; PROC HPFOREST does not generate automatic interaction terms.