[ADMB Users] comparing highly parameterized models

[Fowler, Mark]

Ouch, My apologies. One of the first responses indicated a good chance
that Ian Stewart would reply, and I emulated Pavlov's favourite subjects
when an 'Ian' replied. BTW I got Steve Martell's code working in my
model (DIC and estimated numbers of parameters) so problem solved.

 

Mark Fowler 
Population Ecology Division 
Bedford Inst of Oceanography 
Dept Fisheries & Oceans 
Dartmouth NS Canada 
B2Y 4A2 
Tel. (902) 426-3529 
Fax (902) 426-9710 
Email Mark.Fowler at dfo-mpo.gc.ca 
Home Tel. (902) 461-0708 
Home Email mark.fowler at ns.sympatico.ca 

From: Ian Taylor - NOAA Federal [mailto:ian.taylor at noaa.gov] 
Sent: May 3, 2013 1:43 PM
To: Fowler, Mark
Cc: John Sibert; users at admb-project.org
Subject: Re: [ADMB Users] comparing highly parameterized models

 

Credit where credit is due: it was Ian Stewart, not me, who developed
the approach and responded to the previous emails.

-Ian

 

On Fri, May 3, 2013 at 5:04 AM, Fowler, Mark <Mark.Fowler at dfo-mpo.gc.ca>
wrote:

Thanks, John, I've got it. Slight misunderstanding, my subject header
was inadequate. I'm okay for comparing models where the number of
parameters are easily determined. My question concerned determination of
the number of 'real' parameters represented by deviation parameters. Ian
Taylor developed an approach to approximate the number of parameters for
comparing Stock Synthesis models (the excerpt in my email), and I was
asking for details on how a couple of scalars used by Ian were derived.
Both Ian and Steve Martell responded, giving me a handle on them. I'll
need to compute DIC's to estimate the effective number of parameters.


>       Mark Fowler
                Population Ecology Division
>       Bedford Inst of Oceanography
>       Dept Fisheries & Oceans
>       Dartmouth NS Canada
                B2Y 4A2
                Tel. (902) 426-3529 <tel:%28902%29%20426-3529> 
                Fax (902) 426-9710 <tel:%28902%29%20426-9710> 
                Email Mark.Fowler at dfo-mpo.gc.ca
                Home Tel. (902) 461-0708 <tel:%28902%29%20461-0708> 
                Home Email mark.fowler at ns.sympatico.ca



-----Original Message-----
From: users-bounces at admb-project.org
[mailto:users-bounces at admb-project.org] On Behalf Of John Sibert
Sent: May 2, 2013 4:03 PM
To: users at admb-project.org
Subject: Re: [ADMB Users] comparing highly parameterized models

A less garbled link perhaps
http://www.soest.hawaii.edu/PFRP/reprints/multifan.pdf

John Sibert
Emeritus Researcher, SOEST
University of Hawaii at Manoa
Honolulu HI (GMT-10)
808-294-3842

Visit the ADMB project http://admb-project.org/

On 05/02/2013 08:15 AM, dave fournier wrote:
>
> This is far too  long to read, but we approached model selection in
> the multifan cl model using approximate Bayes factors whihc enabled us

> to copmoare non nested models.
>
> This link might have survived being copied.
>
>
>
> http://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&
> sqi=2&ved=0CEAQFjAD&url=http%3A%2F%2Fwww.soest.hawaii.edu%2Fpfrp%2Frep
> rints%2Fmultifan.pdf&ei=lKyCUc-lMYPHiwKk_YDQDA&usg=AFQjCNEW-tR28NpujfA
> kDNC5BDVIumdQhA&sig2=49idpQ7NXkxhrLbIYcULfg&bvm=bv.45960087,d.cGE
>
> _______________________________________________
> Users mailing list
> Users at admb-project.org
> http://lists.admb-project.org/mailman/listinfo/users
>

_______________________________________________
Users mailing list
Users at admb-project.org
http://lists.admb-project.org/mailman/listinfo/users
_______________________________________________
Users mailing list
Users at admb-project.org
http://lists.admb-project.org/mailman/listinfo/users

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.admb-project.org/pipermail/users/attachments/20130503/5548e881/attachment.html>