CP4SBI: Local Conformal Calibration of Credible Sets in Simulation-Based Inference

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cabezas, Luben M. C., Santos, Vagner S., Ramos, Thiago R., Rodrigues, Pedro L. C., Izbicki, Rafael
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914008358977536
author Cabezas, Luben M. C.
Santos, Vagner S.
Ramos, Thiago R.
Rodrigues, Pedro L. C.
Izbicki, Rafael
author_facet Cabezas, Luben M. C.
Santos, Vagner S.
Ramos, Thiago R.
Rodrigues, Pedro L. C.
Izbicki, Rafael
contents Current experimental scientists have been increasingly relying on simulation-based inference (SBI) to invert complex non-linear models with intractable likelihoods. However, posterior approximations obtained with SBI are often miscalibrated, causing credible regions to undercover true parameters. We develop $\texttt{CP4SBI}$, a model-agnostic conformal calibration framework that constructs credible sets with local Bayesian coverage. Our two proposed variants, namely local calibration via regression trees and CDF-based calibration, enable finite-sample local coverage guarantees for any scoring function, including HPD, symmetric, and quantile-based regions. Experiments on widely used SBI benchmarks demonstrate that our approach improves the quality of uncertainty quantification for neural posterior estimators using both normalizing flows and score-diffusion modeling.
format Preprint
id arxiv_https___arxiv_org_abs_2508_17077
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CP4SBI: Local Conformal Calibration of Credible Sets in Simulation-Based Inference
Cabezas, Luben M. C.
Santos, Vagner S.
Ramos, Thiago R.
Rodrigues, Pedro L. C.
Izbicki, Rafael
Machine Learning
Current experimental scientists have been increasingly relying on simulation-based inference (SBI) to invert complex non-linear models with intractable likelihoods. However, posterior approximations obtained with SBI are often miscalibrated, causing credible regions to undercover true parameters. We develop $\texttt{CP4SBI}$, a model-agnostic conformal calibration framework that constructs credible sets with local Bayesian coverage. Our two proposed variants, namely local calibration via regression trees and CDF-based calibration, enable finite-sample local coverage guarantees for any scoring function, including HPD, symmetric, and quantile-based regions. Experiments on widely used SBI benchmarks demonstrate that our approach improves the quality of uncertainty quantification for neural posterior estimators using both normalizing flows and score-diffusion modeling.
title CP4SBI: Local Conformal Calibration of Credible Sets in Simulation-Based Inference
topic Machine Learning
url https://arxiv.org/abs/2508.17077