Corpus ID: 221954280

Finite mixture models do not reliably learn the number of components

@inproceedings{Cai2021FiniteMM,
  title={Finite mixture models do not reliably learn the number of components},
  author={Diana Cai and Trevor Campbell and Tamara Broderick},
  booktitle={ICML},
  year={2021}
}
Scientists and engineers are often interested in learning the number of subpopulations (or components) present in a data set. A common suggestion is to use a finite mixture model (FMM) with a prior on the number of components. Past work has shown the resulting FMM component-count posterior is consistent; that is, the posterior concentrates on the true generating number of components. But existing results crucially depend on the assumption that the component likelihoods are perfectly specified… Expand

Figures from this paper

References

SHOWING 1-10 OF 86 REFERENCES
Convergence of latent mixing measures in finite and infinite mixture models
Finite Mixture and Markov Switching Models
On the Identifiability of Finite Mixtures
A simple example of Dirichlet process mixture inconsistency for the number of components
PROBABILITY AND MEASURE
...
1
2
3
4
5
...