HP Labs Technical Reports
Click here for full text:
A Large Deviations Principle for Dirichlet Process Posteriors
Ganesh, A.J.; O'Connell, Neil
Keyword(s): large deviations; Bayesian statistics; Dirichlet process
Abstract: Let X(sub k) be a sequence of iid random variables taking values in a compact metric space omega, and consider the problem of estimating the law of X(sub 1) in a Bayesian framework. A conjugate family of priors for non- parametric Bayesian inference are the Dirichlet process priors popularized by Ferguson. We prove that if the prior distribution is Dirichlet, then the sequence of posterior distributions satisfies a large deviation principle, and give an explicit expression for the rate function.
Back to Index