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

Posted to dev@commons.apache.org by "Becksfort, Jared" <Ja...@STJUDE.ORG> on 2013/03/19 15:32:09 UTC

[math] Expectation Maximization Code Location

Hello,

I recently submitted some code for the fitting of multivariate normal mixture model distributions to data using the Expectation Maximization algorithm.  I placed the code in the o.a.c.m.optimization.fitting package.  Gilles wonders if it would make more sense to have it in the o.a.c.m.distribution.fitting package.  That may be true.

Any thoughts?  Either one is fine by me.  EM is an optimization algorithm and it also fits distributions.  It uses a lot of the code from o.a.c.m.distribution and none from o.a.c.m.optimization, so that may imply that it would make sense to move it.

The code is not yet committed.  Here is the link: https://issues.apache.org/jira/browse/MATH-817

Thanks,
Jared

________________________________
Email Disclaimer: www.stjude.org/emaildisclaimer
Consultation Disclaimer: www.stjude.org/consultationdisclaimer

Re: [math] Expectation Maximization Code Location

Posted by Luc Maisonobe <Lu...@free.fr>.

Le 19/03/2013 15:32, Becksfort, Jared a écrit :
> Hello,
> 
> I recently submitted some code for the fitting of multivariate normal mixture model distributions to data using the Expectation Maximization algorithm.  I placed the code in the o.a.c.m.optimization.fitting package.  Gilles wonders if it would make more sense to have it in the o.a.c.m.distribution.fitting package.  That may be true.
> 
> Any thoughts?  Either one is fine by me.  EM is an optimization algorithm and it also fits distributions.  It uses a lot of the code from o.a.c.m.distribution and none from o.a.c.m.optimization, so that may imply that it would make sense to move it.
> 
> The code is not yet committed.  Here is the link: https://issues.apache.org/jira/browse/MATH-817

I am not a specialist in these area, but also feel EM would be more at
home with distributions er even statistics than in optimization.

best regards,

Luc

> 
> Thanks,
> Jared
> 
> ________________________________
> Email Disclaimer: www.stjude.org/emaildisclaimer
> Consultation Disclaimer: www.stjude.org/consultationdisclaimer
> 


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org