Solving Inverse Parametrized Problems via Finite Elements and Extreme Learning Networks

TL;DR

This paper introduces a new interpolation-based framework combining finite element methods and extreme learning networks for solving parameter-dependent PDEs, with applications to inverse problems like photoacoustic tomography.

Contribution

It develops a unified approach that handles low- and high-dimensional parameter spaces, providing error estimates and computational advantages over traditional methods.

Findings

Error bounds are derived for both low- and high-dimensional parameter spaces.

The framework achieves substantial computational savings in inverse problem applications.

Rigorous error estimates link spatial discretization and parameter approximation.

Abstract

We develop an interpolation-based modeling framework for parameter-dependent partial differential equations arising in control, inverse problems, and uncertainty quantification. The solution is discretized in the physical domain using finite element methods, while the dependence on a finite-dimensional parameter is approximated separately. We establish existence, uniqueness, and regularity of the parametric solution and derive rigorous error estimates that explicitly quantify the interplay between spatial discretization and parameter approximation. In low-dimensional parameter spaces, classical interpolation schemes yield algebraic convergence rates based on Sobolev regularity in the parameter variable. In higher-dimensional parameter spaces, we replace classical interpolation by extreme learning machine (ELM) surrogates and obtain error bounds under explicit approximation and stability assumptions. The proposed framework is applied to inverse problems in quantitative photoacoustic tomography, where we derive potential and parameter reconstruction error estimates and demonstrate substantial computational savings compared to standard approaches, without sacrificing accuracy.