[Corpora-List] Significance tests for n-fold cross validation

[Angus B. Grieve-Smith]

On 5/8/2011 3:28 PM, Emad Mohamed wrote:
> My question is: Do I need to use any significance test to show  that 
> adding feature x really helps? if no, why not? and if yes, what are 
> the appropriate tests in such a case?

     Significance tests are only relevant if you're testing a 
representative sample of the data.  If your sample is not 
representative, no amount of significance testing will make it so.

-- 
				-Angus B. Grieve-Smith
				Saint John's University
				grvsmth at panix.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20110508/84900322/attachment.htm>
-------------- next part --------------
_______________________________________________
UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora