I am working with mixture models. I fitted more than one model to the data and then try to select the most appropriate model using different selection criteria, for example, AIC. My supervisors asked me to use different comparison criteria such as Vuong test. Vuong test is a likelihood ratio test for non-nested models. I have read several papers about mixture models, and almost all of them used selection information criteria as comparison criteria. I did not find any paper that used Vuong test in a mixture context. Is that mean I cannot use Vuong test for mixture models?

Any help, please?

  • $\begingroup$ Maybe no one knows about it? I don't. But I know AIC and BIC and that cross validation is better than any other test I've seen. $\endgroup$ – keiv.fly Oct 21 '18 at 8:03
  • $\begingroup$ Roughly speaking, trying to find the model that fits the data better isn't the same thing as testing a model by assessing the discrepancy of the data with it. So it depends which you want to do. $\endgroup$ – Scortchi Nov 8 '18 at 9:23
  • $\begingroup$ @Scortchi thank you so much for your comment. I would like to use Vuong test for comparing two different models and to see which of them fits the data better than the other. $\endgroup$ – Maryam Nov 8 '18 at 9:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.