[jira] Commented: (MAHOUT-29) Gibb's sampling implementation

["Ted Dunning (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/MAHOUT-29?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12586848#action_12586848 ] 

Ted Dunning commented on MAHOUT-29:
-----------------------------------


The umacs package in R provides a very nice implementation of an integrated Gibbs and Metropolis framework.

http://cran.r-project.org/web/packages/Umacs/index.html



> Gibb's sampling implementation
> ------------------------------
>
>                 Key: MAHOUT-29
>                 URL: https://issues.apache.org/jira/browse/MAHOUT-29
>             Project: Mahout
>          Issue Type: New Feature
>            Reporter: Isabel Drost
>
> Copied over from original issue:
> > EM clustering can also be changed very slightly by assigning points to single clusters chosen at random according to the probability of membership. This
> > turns EM clustering into Gibb's sampling. The important property that is changed is that you now can sample over the distribution of possible samplings
> > which can be very important if some parts of your data are well defined and some parts not so well defined.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.