- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Hello Everyone,
We have an application score model built in SAS E-miner. And for the production data, we wish to estimate the important measures C-Stat and KS statistics on the production data for monitoring purposes. Wish to know some guidance and an easy way to estimate these. Can someone help on this?
Thanks
Kind regards,
- Tags:
- monitoring
- SAS Miner
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
It must be Credit Scorecard .
data have; set sashelp.class(keep=sex height rename=(sex=good_bad height=total_score)); run; proc logistic data=have ; model good_bad(event='M')=total_score/nofit; roc 'Test ROC' pred=total_score; run; /********KS**********/ proc npar1way data=have edf ; class good_bad; var total_score; run;
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Hi, Thanks for your help. I see that you train the dataset and see the ROC and KS. In my case, kindly note that I am already having a model in SAS E-Miner (Scorecard) and wish to add a new Test dataset for scoring to have fit statistics including ROC/KS. This is basically to monitor the scorecard model for the production data over the period. Thanks so much for taking your time
Kind regards,
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
There is a SCORE node . connent your scorecard node and test dataset node to it .
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Thank you so much. However, the score node does not provide the Fit Statistics. I see few ways using the following threads and will try that. Thanks a lot.
Kind regards,