Proc Dtree VS Proc Hpsplit

YuTing Tian
Independent


Abstract

Regression and classification trees are methods for analyzing how a dependent variable is correlated with independent variables. Proc Hpsplit and Proc Dtree can both create decision trees that look similar. Both begin with a single node followed by increased number of leaves. However, they focus on a different purpose. This paper is a preliminary introduction differences between Proc Hpsplit and Proc Dtree.