https://communities.sas.com/t5/SAS-Data-Science/Scoring-with-Upper-Confidence-Limit-Estimate/m-p/248118#M3656 <P>I've used the HPGLM node to build a regression and have the upper confidcence limits from the parameter table, but am looking to understand if there is a way to use them to create the Score Code instead of the standard estimates.</P><P>&nbsp;</P><P>I am already committed to re-building my regression with PROC HPGENSELECT by hand, so I can change ALPHA.&nbsp; It would be great to have both of these as optional parameters within E-Miner.</P> Thu, 04 Feb 2016 22:29:07 GMT lapomarda 2016-02-04T22:29:07Z https://communities.sas.com/t5/SAS-Data-Science/Scoring-with-Upper-Confidence-Limit-Estimate/m-p/248118#M3656 <P>I've used the HPGLM node to build a regression and have the upper confidcence limits from the parameter table, but am looking to understand if there is a way to use them to create the Score Code instead of the standard estimates.</P><P>&nbsp;</P><P>I am already committed to re-building my regression with PROC HPGENSELECT by hand, so I can change ALPHA.&nbsp; It would be great to have both of these as optional parameters within E-Miner.</P> Thu, 04 Feb 2016 22:29:07 GMT https://communities.sas.com/t5/SAS-Data-Science/Scoring-with-Upper-Confidence-Limit-Estimate/m-p/248118#M3656 lapomarda 2016-02-04T22:29:07Z https://communities.sas.com/t5/SAS-Data-Science/Scoring-with-Upper-Confidence-Limit-Estimate/m-p/387019#M5738 <P>Thanks for the feedback! &nbsp;There is currently no way to use the upper confidence limits as the estimates in the score code - I don't know that we've heard that request before, but interesting to hear this! &nbsp;</P> <P>&nbsp;</P> <P>The ability to change ALPHA for a GLM is something we are adding into our future data mining and machine learning product.</P> Thu, 10 Aug 2017 14:27:45 GMT https://communities.sas.com/t5/SAS-Data-Science/Scoring-with-Upper-Confidence-Limit-Estimate/m-p/387019#M5738 WendyCzika 2017-08-10T14:27:45Z