vix.ing · top · new · best · stats · spec

De Finetti + Sanov = Bayes: Exchangeable Prediction under Moment Constraints

2025/09/16 by Nicholas G. Polson, Daniel Zantedeschi, Polson, Nicholas G. +1 · 2 citations
Mathematics · #math.ST #msc:60F10 #msc:60G09 #msc:62F15 #msc:94A17 #stat.ME #stat.TH

paper · pdf · doi:10.48550/arxiv.2509.13283

v3: synchronized with the final manuscript

arxiv created 2026/07/28 · arxiv updated 2026/07/30

Abstract

We study exchangeable prediction when empirical-moment constraints define each active finite horizon N. The relevant law is the de Finetti mixture conditioned on EN = PhatN in EepsN. By permutation invariance, the target may be any fixed block of m coordinates within the active horizon, regardless of whether those coordinates are labeled past, held out, or future relative to any finite cut. Conditionally on the directing measure mu, the Gibbs-conditioning principle sends the law of such a block to the m-fold product of the I-projection P*mu = argminQ in E D(Q || mu). On a finite alphabet we give an elementary master inequality for general polyhedral moment windows. After mixing over the constraint posterior PiN,E, and under weak convergence plus posterior-averaged component control, the finite-dimensional marginals converge to a consistent exchangeable law whose random directing measure is the I-projection P*mu, with mu drawn from the weak limit PiE. Sequential prediction under this limiting law is therefore Bayesian prediction from a mixture of componentwise I-projections. The limiting behavior depends on reachability. For a reachable constraint, the projection is asymptotically the identity on the selected subfamily. Under additional regularity, an unreachable constraint makes the constraint posterior concentrate on the rate-minimizing subfamily, while the projections remain nontrivial. In our examples, at least one operation is asymptotically inactive, though both enter the finite-horizon construction. The master bound also reads as an equivalence of ensembles. We reserve "maximum entropy" for a uniform or flat baseline and use "minimum relative entropy" or "I-projection" in general.

Citations

Cited by

Related