Some cautionary tales about Bayesian predictive inference
Statistics Theory
2026-07-21 v1 Methodology
Machine Learning
Abstract
Two misunderstandings, frequently arising in Bayesian predictive inference, are discussed. The first deals with the data generating mechanism, while the second consists in overestimating the role played by asymptotic exchangeability. Some consequences of such misunderstandings are highlighted through examples.
Cite
@article{arxiv.2607.19206,
title = {Some cautionary tales about Bayesian predictive inference},
author = {Emanuela Dreassi and Fabrizio Leisen and Luca Pratelli and Pietro Rigo},
journal= {arXiv preprint arXiv:2607.19206},
year = {2026}
}