RPM - SAS EG

[andreas_zaras]

Hello,

In RPM of SAS Enterprise Guide the software outputs a scorecard. The characteristics are chosen via a logistic regression when the target variable is e.g. binary and for each attribute scorecard points are given. In the scorecard that is developed what is the the score used and the points to double the odds ratio? score=offset+factor*ln(odds). E.g. in SAS Scorecard nodes of SAS EM and in SAS VDMML the score is 200, the odds ratio is ln(50/1) and the points to double odds are 20. Finally the regression is done on the WOE or the original values of the data set?

Thanks in advance,

Andreas

[andreas_zaras · Accepted solution]

I found the post presented in the link below that has shed some light to my questions.

https://communities.sas.com/t5/SAS-Communities-Library/Tip-How-to-interpret-your-SAS-Rapid-Predictiv...