Combining Decision Trees with Regression in Predictive Modeling with SAS® Enterprise MinerTM

Abstract

The purpose of this paper is to illustrate how the Decision Tree node can be used to optimally bin the inputs for use in a logistic regression. Binning can be viewed as a complex non-linear transformation of the inputs. By utilizing this technique, we may be able to capture complex non-linear relationships between an input variable and the target variable… (More)

Topics

1 Figure or Table

Cite this paper

@inproceedings{Sarma2005CombiningDT, title={Combining Decision Trees with Regression in Predictive Modeling with SAS® Enterprise MinerTM}, author={Kattamuri S. Sarma}, year={2005} }