First-Order Conditions for the Optimal Control of Learning-Informed Nonsmooth PDEs

DOI10.1080/01630563.2025.2488796MaRDI QIDQ6961994zbMATH OpenFDO

Authors Guozhi Dong, Michael Hintermüller, Kostas Papafitsoros, Kathrin Völkner

Publication date 9 July 2025

Published in Numerical Functional Analysis and Optimization (Search for Journal in Brave)

Copyright license Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Full work available at URL https://doi.org/10.1080/01630563.2025.2488796

Mathematics Subject Classification ID

Existence problems for PDEs: global existence, local existence, non-existence (35A01) Uniqueness problems for PDEs: global uniqueness, local uniqueness, non-uniqueness (35A02) Boundary value problems for second-order elliptic equations (35J25) Semilinear elliptic equations (35J61) PDEs in connection with control and optimization (35Q93) PDE constrained optimization (numerical aspects) (49M41) Artificial neural networks and deep learning (68T07) Control/observation systems governed by partial differential equations (93C20)

Review text

This work addresses the optimal control problem subjected to the semi-linear elliptic PDE, where the nonlinear term is determined by complex non-smooth functions, among which the main interest is attracted to the Rectified Linear Unit (ReLU) systems widely used in artificial neural networks. Thus, the problem is reduced to the so-called concept of learning-informed PDE. For this problem, the authors studied the existence and uniqueness of the solution, differentiability, and the approximation properties.

Cites work

• A bundle-free implicit programming approach for a class of elliptic MPECs in function space
• A descent algorithm for the optimal control of ReLU neural network informed PDEs based on approximate directional derivatives
• CONTINUOUS PIECEWISE LINEAR FUNCTIONS
• Deep learning
• Discretization of parameter identification in PDEs using neural networks
• Error bounds for approximations with deep ReLU neural networks in \(W^{s , p}\) norms
• Generalization of Hinging Hyperplanes
• scientific article; zbMATH DE number 5703572 (Why is no real title available?)
• scientific article; zbMATH DE number 3917080 (Why is no real title available?)
• scientific article; zbMATH DE number 3569238 (Why is no real title available?)
• scientific article; zbMATH DE number 1405266 (Why is no real title available?)
• scientific article; zbMATH DE number 2208228 (Why is no real title available?)
• Learning-informed parameter identification in nonlinear time-dependent PDEs
• Optimal approximation of piecewise smooth functions using deep ReLU neural networks
• Optimal Control in Some Variational Inequalities
• Optimal control of a non-smooth semilinear elliptic equation
• Optimization with learning-informed differential equation constraints and its applications
• Optimization with PDE Constraints
• Quantitative Magnetic Resonance Imaging: From Fingerprinting to Integrated Physics-Based Models
• Set-valued analysis
• Several approaches for the derivation of stationarity conditions for elliptic MPECs with upper-level control constraints
• Strong stationarity for a highly nonsmooth optimization problem with control constraints
• Variable step mollifiers and applications

Cited in

(1)

• Stationarity conditions for bilinear optimal control of a non-smooth semilinear elliptic equation

This page was built for publication: First-Order Conditions for the Optimal Control of Learning-Informed Nonsmooth PDEs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6961994)