In Enterprise Miner I used Interactive grouping to binning my input variables. I connected to the unbinned variables the regression node and ran a logistic regression. Then from the Interactive Grouping node I ran a logistic regression and a scorecard.
The results were ROC for the logistic regression without binning was .839; an improvement was seen in the logistic regression run on the binned variables (ROC = .892); the scorecard run off of the binned variables gave an ROC of .888. (see attached image)
I expected the binned variables to give a higher ROC than unbinned, but I'm unsure why the regression and scorecard yielded different results. The Model Comparison node compares the output statistics very conveniently; is there a way to compare the relevant settings between the regression and scorecard.
Thanks in advance for any suggestions. -- George Rezek
If you are just running the Regression and Scorecard nodes after the Interactive Grouping node with defaults, then one thing you might want to check are the inputs used in the Regression node - by default, it is using both the GRP_ and the WOE_ variables for each input whereas the Scorecard node is only using the WOE_ variables. When I set Use=No for the GRP_ variables in the variables editor for the Regression node, then my assessment results matched.
HI @grezek,
Thanks for your question! Perhaps this webinar and community article will help here? Model Selection in SAS Enterprise Guide and SAS Enterprise Miner
Best,
Anna
Thanks, Anna. I'll check it out. -- George Rezek
If you are just running the Regression and Scorecard nodes after the Interactive Grouping node with defaults, then one thing you might want to check are the inputs used in the Regression node - by default, it is using both the GRP_ and the WOE_ variables for each input whereas the Scorecard node is only using the WOE_ variables. When I set Use=No for the GRP_ variables in the variables editor for the Regression node, then my assessment results matched.
Available on demand!
Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.
Use this tutorial as a handy guide to weigh the pros and cons of these commonly used machine learning algorithms.
Find more tutorials on the SAS Users YouTube channel.