Markov chain sampling methods for Dirichlet process mixture

This article reviews Markov chain methods for sampling from the posterior d istribution of a Dirichlet process mixture model and presents two new class es of methods. One new approach is to make Metropolis-Hastings updates of t he indicators specifying which mixture component is associated with each ob servation, perhaps supplemented with a partial form of Gibbs sampling, The other new approach extends Gibbs sampling for these indicators by using a s et of auxiliary parameters. These methods are simple to implement and are m ore efficient than previous ways of handling general Dirichlet process mixt ure models with non-conjugate priors.