Learning-informed parameter identification in nonlinear time-dependent PDEs
From MaRDI portal
Publication:6073849
Abstract: We introduce and analyze a method of learning-informed parameter identification for partial differential equations (PDEs) in an all-at-once framework. The underlying PDE model is formulated in a rather general setting with three unknowns: physical parameter, state and nonlinearity. Inspired by advances in machine learning, we approximate the nonlinearity via a neural network, whose parameters are learned from measurement data. The later is assumed to be given as noisy observations of the unknown state, and both the state and the physical parameters are identified simultaneously with the parameters of the neural network. Moreover, diverging from the classical approach, the proposed all-at-once setting avoids constructing the parameter-to-state map by explicitly handling the state as additional variable. The practical feasibility of the proposed method is confirmed with experiments using two different algorithmic settings: A function-space algorithm based on analytic adjoints as well as a purely discretized setting using standard machine learning algorithms.
Cites work
- A convergence rates result for Tikhonov regularization in Banach spaces with non-smooth operators
- All-at-once versus reduced iterative methods for time dependent inverse problems
- An introduction to the mathematical theory of inverse problems
- Coupled regularization with multiple data discrepancies
- Deep Network Approximation for Smooth Functions
- Design of the monodomain model by artificial neural networks
- Discovering governing equations from data by sparse identification of nonlinear dynamical systems
- Goal oriented adaptivity in the IRGNM for parameter identification in PDEs. II: All-at-once formulations
- scientific article; zbMATH DE number 4194524 (Why is no real title available?)
- scientific article; zbMATH DE number 5703572 (Why is no real title available?)
- scientific article; zbMATH DE number 3576139 (Why is no real title available?)
- scientific article; zbMATH DE number 936298 (Why is no real title available?)
- Integration based profile likelihood calculation for PDE constrained parameter estimation problems
- Iterative regularization methods for nonlinear ill-posed problems
- Landweber–Kaczmarz for parameter identification in time-dependent inverse problems: all-at-once versus reduced version
- Learning partial differential equations via data discovery and sparse optimization
- Local convergence of the steepest descent method in Hilbert spaces
- Multicomposite nonconvex optimization for training deep neural networks
- Multilayer feedforward networks are universal approximators
- Neural network approximation
- Nonlinear hemivariational inequalities with eigenvalues near zero
- Nonlinear optimization.
- Nonlinear partial differential equations with applications
- Optimal control of a non-smooth semilinear elliptic equation
- Optimization with learning-informed differential equation constraints and its applications
- Preconditioned all-at-once methods for large, sparse parameter estimation problems
- Regularization based on all-at-once formulations for inverse problems
- Semigroups of linear operators and applications to partial differential equations
- Solving inverse problems using data-driven models
- The modern mathematics of deep learning
- The tangential cone condition for some coefficient identification model problems in parabolic PDEs
Cited in
(9)- Parameter identification of Bouc–Wen type models using a modified experience-based learning algorithm
- Relaxation approach for learning neural network regularizers for a class of identification problems
- On the identification and optimization of nonsmooth superposition operators in semilinear elliptic PDEs
- Bi-level iterative regularization for inverse problems in nonlinear PDEs
- A descent algorithm for the optimal control of ReLU neural network informed PDEs based on approximate directional derivatives
- First-Order Conditions for the Optimal Control of Learning-Informed Nonsmooth PDEs
- Sequential bi-level regularized inversion with application to hidden reaction law discovery
- The extended adjoint state and nonlinearity in correlation-based passive imaging
- Global convergence of adjoint-optimized neural PDEs
This page was built for publication: Learning-informed parameter identification in nonlinear time-dependent PDEs
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6073849)
