Amortized low-rank approximation for hyperparameter marginalization in PDE-governed Bayesian inverse problems

DOI10.48550/arXiv.2607.03355MaRDI QIDQ7283920zbMATH OpenarXivFDO

Authors Georg Stadler, Sonia Reilly

Publication date 3 July 2026

Published in arXiv

Tldr We propose two scalable methods to efficiently approximate log-determinant ratios and maximum a posteriori estimations, which are required for Bayesian hyperparameter marginalization in PDE-governed inverse problems with nonlinear prior hyperparameters, achieving a factor of 30-45 speedup relative to the direct method.

Full work available at URL https://doi.org/10.48550/arXiv.2607.03355

Mathematics Subject Classification ID

PDEs in connection with statistics (35Q62) Inverse problems for PDEs (35R30) Bayesian inference (62F15) Numerical methods for inverse problems for initial value and initial-boundary value problems involving PDEs (65M32)

arXiv classification

math.NAmath.OC

This page was built for publication: Amortized low-rank approximation for hyperparameter marginalization in PDE-governed Bayesian inverse problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7283920)