As the first step in the decommissioning of sasCommunity.org the site has been converted to read-only mode.


Here are some tips for How to share your SAS knowledge with your professional network.


Difference between revisions of "Tips:Implement Regularized Discriminant Analysis in SAS"

From sasCommunity
Jump to: navigation, search
m (gardening)
 
(4 intermediate revisions by 4 users not shown)
Line 1: Line 1:
 
Friedman proposed Regularized Discriminant Analysis (RDA) to overcome multicollinearity and other causes that make LDA/QDA ill-conditioned. The core idea is to regularize illy-conditioned within class covariance matrix with the pooled covariance matrix for QDA or to regularize illy-conditioned pooled covariance matrix with a diagonal matrix for LDA. For details about this algorithm, check the book: Elements of Statistical Learning, Chapter 4, section 3.1.
 
Friedman proposed Regularized Discriminant Analysis (RDA) to overcome multicollinearity and other causes that make LDA/QDA ill-conditioned. The core idea is to regularize illy-conditioned within class covariance matrix with the pooled covariance matrix for QDA or to regularize illy-conditioned pooled covariance matrix with a diagonal matrix for LDA. For details about this algorithm, check the book: Elements of Statistical Learning, Chapter 4, section 3.1.
  
To implement RDA, we output sufficient statistics using OUTSTAT= in PROC DISCRIM and make appropriete changes to relavant statistics, then use the scoring functionality fo PROC DISCRIM to re-score the data with regularized covariance matrix. See link below for sample code on Regularized LDA.
+
To implement RDA, we output sufficient statistics using OUTSTAT= in PROC DISCRIM and make appropriate changes to relevant statistics, then use the scoring functionality for PROC DISCRIM to re-score the data with regularized covariance matrix. See link below for sample code on Regularized LDA.
  
 
<!-- Insert any appropriate category tags. Note that it is important to put the categorys inside the <noinclude> block so the categories are only applied to the tip page.
 
<!-- Insert any appropriate category tags. Note that it is important to put the categorys inside the <noinclude> block so the categories are only applied to the tip page.
Line 7: Line 7:
 
-->
 
-->
 
<noinclude>
 
<noinclude>
[[Category:Data Mining, PROC DISCRIM]]
+
[[Category:Data Mining]]
 +
[[Category:DISCRIM Procedure]]
 
</noinclude>
 
</noinclude>
  
<!-- Provide links to other resources:
 
- If the links are pages on sasCommunity update the "WikiReadMore" line and replace "PAGENAME" with the name of the sasCommunity.org page.
 
- If the links are external (e.g., the sas support site, other external links, and so on, updated the "External ReadMore" line and replace "URL" with the full url including the http:
 
- You can include as many of these lines as you want
 
- if you don't have such links, just delete the line(s)
 
-->
 
 
{{ExternalReadMore|http://www.sas-programming.com/2011/04/regularized-discriminant-analysis.html}}
 
{{ExternalReadMore|http://www.sas-programming.com/2011/04/regularized-discriminant-analysis.html}}
{{WikiReadMore|PAGENAME}}
 
  
 
<!-- Please do not edit below this line, EXCEPT when promoting a tip -->
 
<!-- Please do not edit below this line, EXCEPT when promoting a tip -->
 
<div style="float:right">Submitted By [[User:http&#58;//sas-programming.blospot.com|Liang Xie]]</div>
 
<div style="float:right">Submitted By [[User:http&#58;//sas-programming.blospot.com|Liang Xie]]</div>
 +
 
<noinclude>
 
<noinclude>
[[Category:Tip to be Reviewed]]
+
[[Category:Tip in Use]]
 
</noinclude>
 
</noinclude>

Latest revision as of 17:40, 27 January 2017

Friedman proposed Regularized Discriminant Analysis (RDA) to overcome multicollinearity and other causes that make LDA/QDA ill-conditioned. The core idea is to regularize illy-conditioned within class covariance matrix with the pooled covariance matrix for QDA or to regularize illy-conditioned pooled covariance matrix with a diagonal matrix for LDA. For details about this algorithm, check the book: Elements of Statistical Learning, Chapter 4, section 3.1.

To implement RDA, we output sufficient statistics using OUTSTAT= in PROC DISCRIM and make appropriate changes to relevant statistics, then use the scoring functionality for PROC DISCRIM to re-score the data with regularized covariance matrix. See link below for sample code on Regularized LDA.


....see also


Submitted By Liang Xie