A data scientist specifies PROC HPSPLIT with a GROW ENTROPY option. What splitting criterion does this select?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — GROW ENTROPY picks the information gain criterion, so you're choosing the split that most reduces Shannon entropy. The Gini impurity option is the default with GROW GINI, not entropy.
Full Explanation
The information gain based on Shannon entropy reduction option is correct because GROW ENTROPY in PROC HPSPLIT selects the entropy splitting criterion. It uses Shannon entropy to measure impurity and chooses the split that maximally reduces entropy, which is equivalent to maximizing information gain. The Gini impurity option is not correct because it is the default criterion selected with GROW GINI. The chi-square test statistic option is not correct because it is used in CHAID-style trees with GROW CHISQ. The variance reduction option is not correct because it is used for regression targets and is not invoked by the ENTROPY option.