Partial exchangeability of the prior via shuffling

preprint2014arXivOpen access

Abstract

In inference problems involving a multi-dimensional parameter θθ, it is often natural to consider decision rules that have a risk which is invariant under some group GG of permutations of θθ. We show that this implies that the Bayes risk of the rule is {\em as if} the prior distribution of the parameter is partially exchangeable with respect to GG. We provide a symmetrization technique for incorporating partial exchangeability of θθ into a statistical model, without assuming any other prior information. We refer to this technique as {\em shuffling}. Shuffling can be viewed as an instance of empirical Bayes, where we estimate the (unordered) multiset of parameter values {θ1,θ2,…,θp}\{θ_1,θ_2,\dots,θ_p\} while using a uniform prior on GG for their ordering. Estimation of the multiset is a missing data problem which can be tackled with a stochastic EM algorithm. We show that in the special case of estimating the mean-value parameter in a regular exponential family model, shuffling leads to an estimator that is a weighted average of permuted versions of the usual maximum likelihood estimator. This is a novel form of shrinkage.

Explore this paper’s authors, topics and related work.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Reviews 0

Write a reviewWrite

No reviews yet.

Discussion 0

Add a commentWrite

No comments yet.