
Data defines the model by dint of genetic programming, producing the best decile table.


Multinomial Logistic Regression: A Modelfree Approach Bruce Ratner, Ph.D. 

The binary logistic regression is a popular technique for classifying individuals into two mutually exclusive and exhaustive categories, namely, when the target variable is binary: for example, buynot buy, or respondernonresponder. The multinomial logistic regression is the appropriate method when the target variable assumes K (greater than 2) unordered categorical values: for example, read, white, and blue. The multinomial logistic regression model is an assumptionfull, parametric model in which the model structure (equation form) is prespecified. The purpose of this article is to present the Genetic Multinomial Logistic Regression (GOMR) as an assumptionfree, nonparametric model – modelfree where the data defines the predictor variables and the K1 model equations – based on the machine learning paradigm of genetic programming. Pointedly, the multinomial logistic regression’s untenable restriction of model equations having only one set of coefficients across the K1 equations is a nonissue in GMLR. Moreover, the GMLR determines the best set of predictor variables based on a simultaneous and virtually unbiased assessment of all variables, an achievement not possible with the statistical multinomial logistic regression. GMLR is a straightforward extension of the GenIQ Model©, which serves as the modelfree alternative to the binary logistic regression. See companion article A Genetic Logistic Regression Model: A Modelfree Approach to Identifying Responders to a CRM Solicitation

For more information about this article, call Bruce Ratner at 516.791.3544 or 1 800 DM STAT1; or email at br@dmstat1.com. 
Signup for a free GenIQ webcast: Click here. 

