Martin, Ryan
An approximate Bayesian marginal likelihood
approach for estimating nite mixtures
Estimation of nite mixture models when the mixing distribution support is
unknown is an important problem. This paper gives a new approach based on a
marginal likelihood for the unknown support. Motivated by a Bayesian Dirich-
let prior model, a computationally e cient stochastic approximation version of the
marginal likelihood is proposed and large-sample theory is presented. By restricting
the support to a nite grid, a simulated annealing method is employed to maximize
the marginal likelihood and estimate the support. Real and simulated data exam-
ples show that this novel stochastic approximation{simulated annealing procedure
compares favorably to existing methods.
Dirichlet distribution;mixture complexity;predictive recursion
2014-08-07
https://indigo.uic.edu/articles/journal_contribution/An_approximate_Bayesian_marginal_likelihood_approach_for_estimating_nite_mixtures/10771949