turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Data Mining
- /
- Fit Statistics for Rankings in Logistic Regression...

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

03-17-2013 07:50 PM

Hello,

In a (logistic) regression node i don't see the concordance/ discordance fit statistc in the fit statistics window of the results output. Can i find it somewhere else? If this statistic in not produced how are we supposed to check the fit of a logistc regression model that produces rankings (logit scores)?

Thanks in advance,

Andreas

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

Posted in reply to andreas_zaras

08-15-2017 05:46 PM

Since SAS Enterprise Miner was designed for very large data sets which often are fraught with problems such as missing values, there are several classical statistical techniques that are not calculated in SAS Enterprise Miner. In some cases, it is just not possible to compute those metrics for the number of observations in a typical data mining data set.

Instead of relying on those metrics, you can usually validate your model on holdout data in data mining. Candidate models are trained on a portion of the data and the best model can be chosen based on its performance on the holdout data. This would not work for classical methods for which data is often quite limited.

I hope this helps!

Doug