Latent Inversion of Material Coefficients from Boundary Data via Finite Tests and Neural Surrogates
Abstract
We reconstruct a spatially varying material coefficient in a scalar elliptic equation from finitely many boundary excitations, each producing a full Dirichlet trace. To mitigate the ill-posedness and the cost of repeated PDE solves, we restrict the coefficient to a low-dimensional family, specified analytically or learned from samples, and solve the inverse problem in its latent coordinates. We consider a C1 parametrization with m latent coordinates and full-rank derivative at a reference point. If the continuous linearized Neumann-to-Dirichlet map is injective on the corresponding tangent space, at most m excitations suffice for local injectivity and Lipschitz stability. Convergence of the coefficient sensitivities then transfers this stability to conforming finite element discretizations. For sufficiently fine meshes, the stability constant and neighborhood can be chosen independently of the mesh size. Under a local residual-comparison condition, uniform accuracy of the surrogate forward map yields coefficient-error bounds separating representation error, data noise, finite element error, and surrogate error. Derivative accuracy additionally preserves the surrogate's own local stability. All stability statements are local to a reference coefficient. Two-dimensional numerical experiments combine analytic and learned representations of inclusions and crack-like coefficients with neural forward surrogates. They illustrate latent-space reconstruction, reduced online cost, and further improvement from optional FEM-based refinement.
Related
- Efficient Deep Gaussian Process Surrogates via Mini-Batch Latent MAP Estimation
- Causal Mediation Analysis with High-Dimensional Surrogates: Identification, Estimation and Inference under Latent Factor Models
- No Free Lunch in Flow Surrogates under Time-Varying Boundary Conditions: A Two-Regime Study
- Initial and initial boundary value problems for the Navier-Stokes equations with temperature-dependent transport coefficients and large data
- Input-to-State Stabilization of a Coupled ODE-PDE System with Time-Varying Coefficients via Composite Boundary Control
BibTeX
@misc{burman2026latent,
title = {{Latent Inversion of Material Coefficients from Boundary Data via Finite Tests and Neural Surrogates}},
author = {Erik Burman and Mats G. Larson and Karl Larsson and Jonatan Vallin},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.17137},
eprint = {2609.17137},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.17137},
}
RIS
TY - UNPB
TI - Latent Inversion of Material Coefficients from Boundary Data via Finite Tests and Neural Surrogates
AU - Erik Burman
AU - Mats G. Larson
AU - Karl Larsson
AU - Jonatan Vallin
PY - 2026
DA - 2026/09/15
PB - arXiv
DO - 10.48550/arxiv.2609.17137
UR - https://arxiv.org/abs/2609.17137
AB - We reconstruct a spatially varying material coefficient in a scalar elliptic equation from finitely many boundary excitations, each producing a full Dirichlet trace. To mitigate the ill-posedness and the cost of repeated PDE solves, we restrict the coefficient to a low-dimensional family, specified analytically or learned from samples, and solve the inverse problem in its latent coordinates. We consider a C^1 parametrization with m latent coordinates and full-rank derivative at a reference point. If the continuous linearized Neumann-to-Dirichlet map is injective on the corresponding tangent space, at most m excitations suffice for local injectivity and Lipschitz stability. Convergence of the coefficient sensitivities then transfers this stability to conforming finite element discretizations. For sufficiently fine meshes, the stability constant and neighborhood can be chosen independently of the mesh size. Under a local residual-comparison condition, uniform accuracy of the surrogate forward map yields coefficient-error bounds separating representation error, data noise, finite element error, and surrogate error. Derivative accuracy additionally preserves the surrogate's own local stability. All stability statements are local to a reference coefficient. Two-dimensional numerical experiments combine analytic and learned representations of inclusions and crack-like coefficients with neural forward surrogates. They illustrate latent-space reconstruction, reduced online cost, and further improvement from optional FEM-based refinement.
ER -