A Variable Selection Method that Provides a Unique Ranking of Variable Importance
Bruce Ratner, Ph. D.
The unfailing question upon discussing the building of a model: Which are the important candidate predictor variables in rank order? The purpose of this article is to present an unparalleled feature of the GenIQ Model©. GenIQ provides a unique ranking of the candidate predictor variables: The ranking of the relationship between each candidate predictor variable with the target variable – accounting for all candidate predictor variables jointly considered. This is in stark contrast to the statistical correlation coefficient, which provides the ranking of the linear-relationship between each candidate predictor variable with the target variable – without considering any of the other candidate predictor variables. I illustrate the GenIQ feature with several examples comparing the oftentime misleading regression predictor variable coefficients, and the GenIQ Variable Importance Ranking. For an eye-opening preview of the 9-step modeling process of GenIQ, click here.
Related Articles: Data Mining and Its Aplications
1 800 DM STAT-1, or e-mail at firstname.lastname@example.org.