[Corpora-List] robust statistics

Rainer Ottmueller efidetum at googlemail.com
Fri Mar 26 17:27:56 CET 2010


JB> Lots of use of Support Vector Machines Certainly, there will be a return of the investment.

JB> arguably Indeed.


> SVMs are arguably quite consistent with the ideas of robust statistics,
since they are fairly resistant to outliers and certain kinds of model violations. But now, back on topic.

Rainer

On 26 March 2010 16:30, John Burger <john at mitre.org> wrote:


> Rainer Ottmueller wrote:
>
> My first impression is that Robust statistics is not used in all NLP.
>> Please confirm or decline.
>>
>
> Lots of use of Support Vector Machines, especially in topic identification,
> etc. SVMs are arguably quite consistent with the ideas of robust
> statistics, since they are fairly resistant to outliers and certain kinds of
> model violations.


> - John D. Burger
> MITRE
>
>
>
> _______________________________________________
> Corpora mailing list
> Corpora at uib.no
> http://mailman.uib.no/listinfo/corpora
>
-------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/html Size: 3822 bytes Desc: not available URL: <http://www.uib.no/mailman/public/corpora/attachments/20100326/7b5db942/attachment.txt>



More information about the Corpora mailing list