https://communities.sas.com/t5/SAS-Academy-for-Data-Science/quot-Thresholding-quot-technique-for-collapsing-levels/m-p/650184#M812 <P>Re: Predictive Modeling Using Logistic Regression</P> <P>Thresholding for collapsing levels (p.3-17)<BR />When applying thresholding (page 3.17 of course text), instead of grouping all small levels into a single "OTHER", as an alternative approach, would it not make sense to try to aggregate them with the other existing levels, either based on domain knowledge/similarity in meaning (e.g. for residential status, all levels related to "renting" could be grouped together) and/or proportion of response?</P> Sun, 24 May 2020 14:47:41 GMT pvareschi 2020-05-24T14:47:41Z https://communities.sas.com/t5/SAS-Academy-for-Data-Science/quot-Thresholding-quot-technique-for-collapsing-levels/m-p/650184#M812 <P>Re: Predictive Modeling Using Logistic Regression</P> <P>Thresholding for collapsing levels (p.3-17)<BR />When applying thresholding (page 3.17 of course text), instead of grouping all small levels into a single "OTHER", as an alternative approach, would it not make sense to try to aggregate them with the other existing levels, either based on domain knowledge/similarity in meaning (e.g. for residential status, all levels related to "renting" could be grouped together) and/or proportion of response?</P> Sun, 24 May 2020 14:47:41 GMT https://communities.sas.com/t5/SAS-Academy-for-Data-Science/quot-Thresholding-quot-technique-for-collapsing-levels/m-p/650184#M812 pvareschi 2020-05-24T14:47:41Z https://communities.sas.com/t5/SAS-Academy-for-Data-Science/quot-Thresholding-quot-technique-for-collapsing-levels/m-p/650515#M819 <P>Re: Predictive Modeling Using Logistic Regression</P> <P>Thresholding for collapsing levels (p.3-17)<BR />When applying thresholding (page 3.17 of course text), instead of grouping all small levels into a single "OTHER", as an alternative approach, would it not make sense to try to aggregate them with the other existing levels, either based on domain knowledge/similarity in meaning (e.g. for residential status, all levels related to "renting" could be grouped together) and/or proportion of response?</P> <P><FONT color="#0000FF"><STRONG>My response:</STRONG></FONT></P> <P><FONT color="#0000FF"><STRONG>I agree with your comments that rather than dumping rare levels into other group we could use&nbsp; your business knowledge or tools available in SAS EM (Decision tree node, Variable selection mode) and assign rare levels to other correlated levels.</STRONG></FONT></P> Mon, 25 May 2020 19:51:54 GMT https://communities.sas.com/t5/SAS-Academy-for-Data-Science/quot-Thresholding-quot-technique-for-collapsing-levels/m-p/650515#M819 gcjfernandez 2020-05-25T19:51:54Z