You are viewing a plain text version of this content. The canonical link for it is here.

Posted to user@mahout.apache.org by Neil Ghosh <ne...@gmail.com> on 2010/10/02 08:13:30 UTC

Training/Classification techniques in mahout

Hello,

In the Twenty+Newsgroups example it takes an argument

-type cbayes while running.(for Complementary Naive Bayes )

What are other algorithms mahout support and take as an argument ?

Anybody knows the comparisons among them ?

Which one is accurate ? Does it depend on the nature of classification we
are trying to do ?

Thanks
Neil

-- 
Thanks and Regards
Neil
http://neilghosh.com

Re: Training/Classification techniques in mahout

Posted by Ted Dunning <te...@gmail.com>.

-type bayes is the other option.  If time allows cbayes will probably be
better for most purposes.

See http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.13.8572 for
details on the algorithm and comparisons.

On Fri, Oct 1, 2010 at 11:13 PM, Neil Ghosh <ne...@gmail.com> wrote:

> Hello,
>
> In the Twenty+Newsgroups example it takes an argument
>
> -type cbayes while running.(for Complementary Naive Bayes )
>
> What are other algorithms mahout support and take as an argument ?
>
> Anybody knows the comparisons among them ?
>
> Which one is accurate ? Does it depend on the nature of classification we
> are trying to do ?
>
> Thanks
> Neil
>
> --
> Thanks and Regards
> Neil
> http://neilghosh.com
>