Saved in:
Bibliographic Details
Main Authors: Moss, Guy, Muhle, Leah Sophie, Drews, Reinhard, Macke, Jakob H., Schröder, Cornelius
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.22573
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914158078853120
author Moss, Guy
Muhle, Leah Sophie
Drews, Reinhard
Macke, Jakob H.
Schröder, Cornelius
author_facet Moss, Guy
Muhle, Leah Sophie
Drews, Reinhard
Macke, Jakob H.
Schröder, Cornelius
contents Simulation-based inference (SBI) is an established approach for performing Bayesian inference on scientific simulators. SBI so far works best on low-dimensional parametric models. However, it is difficult to infer function-valued parameters, which frequently occur in disciplines that model spatiotemporal processes such as the climate and earth sciences. Here, we introduce an approach for efficient posterior estimation, using a Fourier Neural Operator (FNO) architecture with a flow matching objective. We show that our approach, FNOPE, can perform inference of function-valued parameters at a fraction of the simulation budget of state of the art methods. In addition, FNOPE supports posterior evaluation at arbitrary discretizations of the domain, as well as simultaneous estimation of vector-valued parameters. We demonstrate the effectiveness of our approach on several benchmark tasks and a challenging spatial inference task from glaciology. FNOPE extends the applicability of SBI methods to new scientific domains by enabling the inference of function-valued parameters.
format Preprint
id arxiv_https___arxiv_org_abs_2505_22573
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FNOPE: Simulation-based inference on function spaces with Fourier Neural Operators
Moss, Guy
Muhle, Leah Sophie
Drews, Reinhard
Macke, Jakob H.
Schröder, Cornelius
Machine Learning
Simulation-based inference (SBI) is an established approach for performing Bayesian inference on scientific simulators. SBI so far works best on low-dimensional parametric models. However, it is difficult to infer function-valued parameters, which frequently occur in disciplines that model spatiotemporal processes such as the climate and earth sciences. Here, we introduce an approach for efficient posterior estimation, using a Fourier Neural Operator (FNO) architecture with a flow matching objective. We show that our approach, FNOPE, can perform inference of function-valued parameters at a fraction of the simulation budget of state of the art methods. In addition, FNOPE supports posterior evaluation at arbitrary discretizations of the domain, as well as simultaneous estimation of vector-valued parameters. We demonstrate the effectiveness of our approach on several benchmark tasks and a challenging spatial inference task from glaciology. FNOPE extends the applicability of SBI methods to new scientific domains by enabling the inference of function-valued parameters.
title FNOPE: Simulation-based inference on function spaces with Fourier Neural Operators
topic Machine Learning
url https://arxiv.org/abs/2505.22573