Variationally mimetic operator networks
From MaRDI portal
Abstract: In recent years operator networks have emerged as promising deep learning tools for approximating the solution to partial differential equations (PDEs). These networks map input functions that describe material properties, forcing functions and boundary data to the solution of a PDE. This work describes a new architecture for operator networks that mimics the form of the numerical solution obtained from an approximate variational or weak formulation of the problem. The application of these ideas to a generic elliptic PDE leads to a variationally mimetic operator network (VarMiON). Like the conventional Deep Operator Network (DeepONet) the VarMiON is also composed of a sub-network that constructs the basis functions for the output and another that constructs the coefficients for these basis functions. However, in contrast to the DeepONet, the architecture of these sub-networks in the VarMiON is precisely determined. An analysis of the error in the VarMiON solution reveals that it contains contributions from the error in the training data, the training error, the quadrature error in sampling input and output functions, and a "covering error" that measures the distance between the test input functions and the nearest functions in the training dataset. It also depends on the stability constants for the exact solution operator and its VarMiON approximation. The application of the VarMiON to a canonical elliptic PDE reveals that for approximately the same number of network parameters, on average the VarMiON incurs smaller errors than a standard DeepONet. Further, its performance is more robust to variations in input functions, the techniques used to sample the input and output functions, the techniques used to construct the basis functions, and the number of input functions.
Recommendations
- On the training and generalization of deep operator networks
- Mesh-informed neural networks for operator learning in finite element spaces
- Exponential Convergence of Deep Operator Networks for Elliptic Partial Differential Equations
- MOD-Net: a machine learning approach via model-operator-data network for solving PDEs
- Error estimates for DeepONets: a deep learning framework in infinite dimensions
Cites work
- scientific article; zbMATH DE number 7626805 (Why is no real title available?)
- A Bayesian level set method for geometric inverse problems
- A physics-informed variational DeepONet for predicting crack path in quasi-brittle materials
- Adaptive Finite Element Methods for Elliptic Problems with Discontinuous Coefficients
- B-PINNs: Bayesian physics-informed neural networks for forward and inverse PDE problems with noisy data
- DeepM\&Mnet for hypersonics: predicting the coupled flow and finite-rate chemistry behind a normal shock using neural-network approximation of operators
- DeepM\&Mnet: inferring the electroconvection multiphysics fields based on operator approximation by neural networks
- Error estimates for DeepONets: a deep learning framework in infinite dimensions
- Estimates on the generalization error of physics-informed neural networks for approximating a class of inverse problems for PDEs
- Extended Physics-Informed Neural Networks (XPINNs): A Generalized Space-Time Domain Decomposition Based Deep Learning Framework for Nonlinear Partial Differential Equations
- Imposing Dirichlet boundary conditions with Nitsche's method and spline-based finite elements
- MIONet: Learning Multiple-Input Operators via Tensor Product
- Parallel physics-informed neural networks via domain decomposition
- Physics-informed neural networks for high-speed flows
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
- Scalable uncertainty quantification for deep operator networks using randomized priors
- Scientific machine learning through physics-informed neural networks: where we are and what's next
- Weak imposition of Dirichlet boundary conditions in fluid mechanics
- Über ein Variationsprinzip zur Lösung von Dirichlet-Problemen bei Verwendung von Teilräumen, die keinen Randbedingungen unterworfen sind. (On a variational principle for solving Dirichlet problems less boundary conditions using subspaces)
Cited in
(8)- Mesh-informed neural networks for operator learning in finite element spaces
- MODNO: multi-operator learning with distributed neural operators
- The parametric complexity of operator learning
- Learning WENO for entropy stable schemes to solve conservation laws
- Neural Green's operators for parametric partial differential equations
- Neural dynamical operator: continuous spatial-temporal model with gradient-based and derivative-free optimization methods
- Physics-informed neural operators for efficient modeling of infiltration in porous media
- Stabilizing and solving unique continuation problems by parameterizing data and learning finite element solution operators
This page was built for publication: Variationally mimetic operator networks
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6185143)
