Pseudo-marginal Metropolis–Hastings algorithm

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search

In computational statistics, the pseudo-marginal Metropolis–Hastings algorithm[1] is a Monte Carlo method to sample from a probability distribution. It is an instance of the popular Metropolis–Hastings algorithm that extends its use to cases where the target density is not available analytically. It relies on the fact that the Metropolis–Hastings algorithm can still sample from the correct target distribution if the target density in the acceptance ratio is replaced by an estimate. It is especially popular in Bayesian statistics, where it is applied if the likelihood function is not tractable (see example below).

Algorithm description[edit]

The aim is to simulate from some probability density function . The algorithm follows the same steps as the standard Metropolis–Hastings algorithm except that the evaluation of the target density is replaced by a non-negative and unbiased estimate. For comparison, the main steps of a Metropolis–Hastings algorithm are outlined below.

Metropolis–Hastings algorithm[edit]

Given a current state the Metropolis–Hastings algorithm proposes a new state according to some density . The algorithm then sets with probability

otherwise the old state is kept, that is, .

Pseudo-marginal Metropolis–Hastings algorithm[edit]

If the density is not available analytically the above algorithm cannot be employed. The pseudo-marginal Metropolis–Hastings algorithm in contrast only assumes the existence of an estimator with Now, given and the respective estimate the algorithm proposes a new state according to some density . Next, compute an estimate and set with probability

otherwise the old state is kept, that is, .

Application to Bayesian statistics[edit]

In Bayesian statistics the target of inference is the posterior distribution

where denotes the likelihood function, is the prior and is the prior predictive distribution. Since there is often no analytic expression of this quantity, one often relies on Monte Carlo methods to sample from the distribution instead. Monte Carlo methods often need the likelihood to be accessible for every parameter value . In some cases, however, the likelihood does not have an analytic expression. An example of such a case is outlined below.

Example: Latent variable model[1][2][edit]

Consider a model consisting of i.i.d. latent real-valued random variables with and suppose one can only observe these variables through some additional noise for some conditional density . (This could be due to measurement error, for instance.) We are interested in Bayesian analysis of this model based on some observed data . Therefore, we introduce some prior distribution on the parameter. In order to compute the posterior distribution

we need to find the likelihood function . The likelihood contribution of any observed data point is then

and the joint likelihood of the observed data is

If the integral on the right-hand side is not analytically available, importance sampling can be used to estimate the likelihood. Introduce an auxiliary distribution such that for all then

is an unbiased estimator of and the joint likelihood can be estimated unbiasedly by


  1. ^ a b Christophe Andrieu and Gareth O. Roberts. "The pseudo-marginal approach for efficient Monte Carlo computations". Annals of Statistics. 37.2: 697–725 – via
  2. ^ "Pseudo-marginal MCMC – Building Intelligent Probabilistic Systems". Retrieved 2018-02-08.