Re: [R] mixture models/latent class regression comparison

Thanks for the reply Christian, > I have never used mmlcr for this, but quite generally when fitting such > models, the likelihood has often very many local optima. This means that the > result of the EM (or a similar) algorithm depends on the initialisation, > which in flexmix (and perhaps also i

Re: [R] mixture models/latent class regression comparison

Dear Carson, I have never used mmlcr for this, but quite generally when fitting such models, the likelihood has often very many local optima. This means that the result of the EM (or a similar) algorithm depends on the initialisation, which in flexmix (and perhaps also in mmlcr) is done in a

[R] mixture models/latent class regression comparison

Dear list, I have been comparing the outputs of two packages for latent class regression, namely 'flexmix', and 'mmlcr'. What I have noticed is that the flexmix package appears to come up with a much better fit than the mmlcr package (based on logLik, AIC, BIC, and visual inspection). Has anyone e