# Should the gmm tag be added?

I would like to suggest adding the gmm tag for growth mixture modeling. This would help when asking questions specific to this area. There are a number of questions on this topic, e.g.:

• I've retagged some of those, but in some the GMM part comes in the answer, not the question, and sometimes not the accepted answer. I'm not sure what consensus is on this, I asked about it once before, but no answer was found: meta.stats.stackexchange.com/q/1525/19879 – Korone Mar 5 '13 at 16:45
• @gung makes an excellent point, but I don't really know enough about GMM to write a tag wiki. Can you write it or give us a summary we can use? – Korone Mar 5 '13 at 19:40
• I wrote the tag wiki and it is now placed in the queue for review. – histelheim Mar 7 '13 at 2:17

I don't think that would be a good idea - "GMM" is a little ambiguous and could mean Gaussian Mixture Model, which is currently tagged gaussian-mixture.

Nothing wrong with adding growth-mixture-model as a tag though. It looks like there already exists growth-model which has been used on growth mixture models in the past.

I've created growth-mixture-model as a new tag. I don't know if it should be as synonym of growth-model or if it is a special case.

• I added some examples to the original question – histelheim Mar 5 '13 at 16:38
• A growth-mixture-model would be a special case of a growth-model. The former tries to identify multiple latent populations in the data, whereas the latter assumes a single population that can be approximated by a single growth function. – histelheim Mar 5 '13 at 16:39
• Yes GMM is a bit of a loaded acronym. It could also mean generalized method of moments. – Andy W Mar 5 '13 at 19:11
• I don't necessarily have a problem with your having added a new tag, but please create at least a tag wiki excerpt to guide users on its application. – gung - Reinstate Monica Mar 5 '13 at 19:30

Argh. A few weeks ago, I disambiguated gmm into generalized-moments (the generalized method of moments was too long for a tag) in econometrics; growth-model from social scientists; and gaussian-mixture from computer scientists and machine learners. Unfortunately, gmm as a tag disappeared, along with the explanation of why this is a poor tag (Corone explained it all). If gmm appears again, I will go and weed it out, for the obvious reasons, but a better way would be to avoid it from the outset.

• No worries - the growth-mixture-model tag has been added instead... – histelheim Mar 7 '13 at 15:49