[R] ICL for betamix()

ChrisR chrissair at gmx.net
Tue May 27 22:25:44 CEST 2014


Hi,
I am running beta mixture regressions using the betamix()  command from the
package 'betareg' (3.0-4). I In order to inform the choice about the number
of latent classes I took a look at the various information criteria (AIC,
BIC, ICL) and learnt that the integrated completed likelihood (ICL) is
viewed as the gold standard in determining the number of classes in mixture
models. I noticed, however, that betamix() does not inheret ICL() from the
flexmix() package. Is there a reason for this? And how can I manually
calculate the ICL for a given betamix() object?

Thanks for input,
Chris



--
View this message in context: http://r.789695.n4.nabble.com/ICL-for-betamix-tp4691328.html
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list