R help - Nov 2007 - Automated Binning for building predictive models

Hello, 
Currently I am using R for building a logistic model using numerical and nominal
data as predictors. Before doing the regression, the predictors are grouped. 
The groups I determine manually by trying to maximize the information value
(which is an indicator for the discriminatory power of the variable) under the
condition that enough data are within each group (approx.  5%).

Is there a method in R available which allows for an automated grouping/binning
for that purpose? I know that e.g. SPSS has such a feature. I think such a
feature is helpful in the modeling process.

Regards,
Pano


      Heute schon einen Blick in die Zukunft von E-Mails wagen?
www.yahoo.de/mail
	[[alternative HTML version deleted]]