[R-sig-ME] Weighted-ML estimates

[Murray Jorgensen]

I wonder if in your situation you might be better off using a completely 
EM approach treating both cluster indicator variables and the random 
effects as the missing data?

Murray Jorgensen

H c wrote:
> Hi,
> I'll briefly describe the situation. An EM algorithm is being implemented in
> order to cluster observations into components of a mixture of mixed effects
> models.  It was our hope to use the lme() or lmer() function in the M-step
> to easily find weighted-ML parameter estimates.  Unfortunately, the mixed
> models are often quite complex including serial correlation structures etc.
> 
> 
> Weighted-ML estimates of certain parameters (such as the correlation
> parameters) does not seem likely in closed form or by using lme()/lmer().
>  Due to the weighted-likelihood function, it does not seem appropriate to
> simply use a weighted-least squares approach.  Are appropriate numeric
> approaches available? Has anyone had experience with mixtures of mixed
> models?
> 
> any help would be greatly appreciated,
> 
> Harlan
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

-- 
Dr Murray Jorgensen      http://www.stats.waikato.ac.nz/Staff/maj.html
Department of Statistics, University of Waikato, Hamilton, New Zealand
Email: maj at waikato.ac.nz    majorgensen at ihug.co.nz      Fax 7 838 4155
Phone  +64 7 838 4773 wk    Home +64 7 825 0441    Mobile 021 139 5862