Problem Note 65762: Gradient Boosting models might produce incorrect scores and fit statistics when an offset is used with an interval target
 |  |  |  |  |
In SAS® Visual Data Mining and Machine Learning, gradient-boosting models might produce incorrect scores. As a result, the fit statistics are also incorrect.
The problem occurs in either of these scenarios.
You run PROC GRADBOOST and all of these conditions are met.
- You have an interval target.
- You specify the OFFSET= option.
- You specify DISTRIBUTION=POISSON or DISTRIBUTION= TWEEDIE.
You run the Gradient Boosting node in Model Studio and all of these conditions are met.
- You have an interval target.
- Your Interval target distribution property is set to Tweedie or Poisson.
- The Role is set to Offset for one of your variables.
There are no errors or warnings to indicate a problem.
Click the Hot Fix tab in this note for a link to instructions about accessing and applying the software update.
Operating System and Release Information
| Product Family | Product | System | Product Release | SAS Release | ||
| Reported | Fixed* | Reported | Fixed* | |||
| SAS System | SAS Visual Data Mining and Machine Learning | Microsoft® Windows® for x64 | 8.3 | Viya | ||
| Linux for x64 | 8.3 | Viya | ||||
Viya on Windows: An update for this issue is available for SAS Viya 3.5. For instructions on how to access and apply software updates, see the Updating Your SAS Viya software section in the SAS Viya 3.5 for Windows Deployment Guide at
http://documentation.sas.com/?softwareId=administration&softwareVersion=3.5&softwareContextId=softwareUpdatesWinViya on Linux: An update for this issue is available for SAS Viya 3.5. For instructions on how to access and apply software updates, see the Updating Your SAS Viya software section in the SAS Viya 3.5 for Linux Deployment Guide at
http://documentation.sas.com/?softwareId=administration&softwareVersion=3.5&softwareContextId=softwareUpdates| Type: | Problem Note |
| Priority: | high |
| Date Modified: | 2020-06-08 12:40:31 |
| Date Created: | 2020-03-26 15:43:41 |