English

On the extendibility of finitely exchangeable probability measures

Probability 2016-12-14 v4 Functional Analysis

Abstract

A length-nn random sequence X1,,XnX_1,\ldots,X_n in a space SS is finitely exchangeable if its distribution is invariant under all n!n! permutations of coordinates. Given N>nN > n, we study the extendibility problem: when is it the case that there is a length-NN exchangeable random sequence Y1,,YNY_1,\ldots, Y_N so that (Y1,,Yn)(Y_1,\ldots,Y_n) has the same distribution as (X1,,Xn)(X_1,\ldots,X_n)? In this paper, we give a necessary and sufficient condition so that, for given nn and NN, the extendibility problem admits a solution. This is done by employing functional-analytic and measure-theoretic arguments that take into account the symmetry. We also address the problem of infinite extendibility. Our results are valid when X1X_1 has a regular distribution in a locally compact Hausdorff space SS. We also revisit the problem of representation of the distribution of a finitely exchangeable sequence.

Keywords

Cite

@article{arxiv.1501.06188,
  title  = {On the extendibility of finitely exchangeable probability measures},
  author = {Takis Konstantopoulos and Linglong Yuan},
  journal= {arXiv preprint arXiv:1501.06188},
  year   = {2016}
}