Simulator-free solution of high-dimensional stochastic elliptic partial differential equations using deep neural networks
From MaRDI portal
Publication:2223019
DOI10.1016/J.JCP.2019.109120zbMATH Open1453.65021arXiv1902.05200OpenAlexW2912649832WikidataQ126770901 ScholiaQ126770901MaRDI QIDQ2223019FDOQ2223019
Authors: Sharmila Karumuri, Rohit K. Tripathy, I. Bilionis, Jitesh Panchal 
Publication date: 28 January 2021
Published in: Journal of Computational Physics (Search for Journal in Brave)
Abstract: Stochastic partial differential equations (SPDEs) are ubiquitous in engineering and computational sciences. The stochasticity arises as a consequence of uncertainty in input parameters, constitutive relations, initial/boundary conditions, etc. Because of these functional uncertainties, the stochastic parameter space is often high-dimensional, requiring hundreds, or even thousands, of parameters to describe it. This poses an insurmountable challenge to response surface modeling since the number of forward model evaluations needed to construct an accurate surrogate grows exponentially with the dimension of the uncertain parameter space; a phenomenon referred to as the extit{curse of dimensionality}. State-of-the-art methods for high-dimensional uncertainty propagation seek to alleviate the curse of dimensionality by performing dimensionality reduction in the uncertain parameter space. However, one still needs to perform forward model evaluations that potentially carry a very high computational burden. We propose a novel methodology for high-dimensional uncertainty propagation of elliptic SPDEs which lifts the requirement for a deterministic forward solver. Our approach is as follows. We parameterize the solution of the elliptic SPDE using a deep residual network (ResNet). In a departure from the traditional squared residual (SR) based loss function for training the ResNet, we introduce a novel physics-informed loss function derived from variational principles. Specifically, our loss function is the expectation of the energy functional of the PDE over the stochastic variables. We demonstrate our solver-free approach through various examples where the elliptic SPDE is subjected to different types of high-dimensional input uncertainties. Also, we solve high-dimensional uncertainty propagation and inverse problems.
Full work available at URL: https://arxiv.org/abs/1902.05200
Recommendations
- Deep UQ: learning deep neural network surrogate models for high dimensional uncertainty quantification
- Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
- Solving parametric PDE problems with artificial neural networks
- Exploring the locally low dimensional structure in solving random elliptic PDEs
- Learning in Modal Space: Solving Time-Dependent Stochastic PDEs Using Physics-Informed Neural Networks
inverse problemsenergy functionalelliptic stochastic partial differential equationsdeep residual networkhigh-dimensional uncertainty propagationphysics-informed loss function
Stochastic partial differential equations (aspects of stochastic analysis) (60H15) Artificial neural networks and deep learning (68T07) PDEs with randomness, stochastic partial differential equations (35R60) Numerical solutions to stochastic differential and integral equations (65C30)
Cites Work
- Title not available (Why is that?)
- Title not available (Why is that?)
- Gaussian processes with built-in dimensionality reduction: applications to high-dimensional uncertainty propagation
- Deep learning
- Optimal Scaling of Discrete Approximations to Langevin Diffusions
- Title not available (Why is that?)
- Title not available (Why is that?)
- Multilevel Monte Carlo methods and applications to elliptic PDEs with random coefficients
- The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations
- Quasi-Monte Carlo finite element methods for a class of elliptic partial differential equations with random coefficients
- Multilayer feedforward networks are universal approximators
- On the convergence of generalized polynomial chaos expansions
- Inverse Problem Theory and Methods for Model Parameter Estimation
- Langevin diffusions and the Metropolis-adjusted Langevin algorithm
- Multi-output local Gaussian process regression: applications to uncertainty quantification
- Multi-output separable Gaussian process: towards an efficient, fully Bayesian paradigm for uncertainty quantification
- Uncertainty Quantification and Polynomial Chaos Techniques in Computational Fluid Dynamics
- Kernel principal component analysis for stochastic input model generation
- Large-scale machine learning with stochastic gradient descent
- Computational partial differential equations. Numerical methods and Diffpack programming
- Active subspace methods in theory and practice: applications to kriging surfaces
- Exploiting active subspaces to quantify uncertainty in the numerical simulation of the hyshot II scramjet
- Introduction to uncertainty quantification
- Uncertainty propagation using infinite mixture of gaussian processes and variational Bayesian inference
- Bayesian deep convolutional encoder-decoder networks for surrogate modeling and uncertainty quantification
- The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems
- Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
- Deep UQ: learning deep neural network surrogate models for high dimensional uncertainty quantification
- Modern Monte Carlo variants for uncertainty quantification in neutron transport
- Accelerating Markov chain Monte Carlo with active subspaces
Cited In (46)
- CEENs: causality-enforced evolutional networks for solving time-dependent partial differential equations
- Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data
- Learning to solve Bayesian inverse problems: an amortized variational inference approach using Gaussian and flow guides
- Uncertainty quantification in scientific machine learning: methods, metrics, and comparisons
- A Variational Neural Network Approach for Glacier Modelling with Nonlinear Rheology
- AMS-Net: Adaptive Multiscale Sparse Neural Network with Interpretable Basis Expansion for Multiphase Flow Problems
- Fractional Chebyshev deep neural network (FCDNN) for solving differential models
- Physics-informed information field theory for modeling physical systems with uncertainty quantification
- PI-VEGAN: Physics Informed Variational Embedding Generative Adversarial Networks for Stochastic Differential Equations
- Multifidelity deep operator networks for data-driven and physics-informed problems
- Physics-informed variational inference for uncertainty quantification of stochastic differential equations
- Bayesian Deep Learning Framework for Uncertainty Quantification in Stochastic Partial Differential Equations
- GINNs: graph-informed neural networks for multiscale physics
- A probabilistic generative model for semi-supervised training of coarse-grained surrogates and enforcing physical constraints through virtual observables
- A derivative-free method for solving elliptic partial differential equations with deep neural networks
- Level Set Learning with Pseudoreversible Neural Networks for Nonlinear Dimension Reduction in Function Approximation
- Energy-informed graph transformer model for solid mechanical analyses
- An overview on deep learning-based approximation methods for partial differential equations
- Pre-training strategy for solving evolution equations based on physics-informed neural networks
- Efficient uncertainty quantification for dynamic subsurface flow with surrogate by theory-guided neural network
- Weak form theory-guided neural network (TgNN-wf) for deep learning of subsurface single- and two-phase flow
- A neural network multigrid solver for the Navier-Stokes equations
- Surrogate and inverse modeling for two-phase flow in porous media via theory-guided convolutional neural network
- The deep learning Galerkin method for the general Stokes equations
- A hybrid data-driven-physics-constrained Gaussian process regression framework with deep kernel for uncertainty quantification
- Solving elliptic equations with Brownian motion: bias reduction and temporal difference learning
- Error estimates and physics informed augmentation of neural networks for thermally coupled incompressible Navier Stokes equations
- DeepParticle: learning invariant measure by a deep neural network minimizing Wasserstein distance on data generated from an interacting particle method
- The novel graph transformer-based surrogate model for learning physical systems
- Artificial neural networks for solving elliptic differential equations with boundary layer
- Wavelet neural operator for solving parametric partial differential equations in computational mechanics problems
- Solving Kolmogorov PDEs without the curse of dimensionality via deep learning and asymptotic expansion with Malliavin calculus
- Stochastic projection based approach for gradient free physics informed learning
- A novel sequential method to train physics informed neural networks for Allen Cahn and Cahn Hilliard equations
- WAN discretization of PDEs: best approximation, stabilization, and essential boundary conditions
- Solution of time‐fractional stochastic nonlinear sine‐Gordon equation via finite difference and meshfree techniques
- A Chebyshev Polynomial Neural Network Solver for Boundary Value Problems of Elliptic Equations
- Modeling the dynamics of PDE systems with physics-constrained deep auto-regressive networks
- On the eigenvector bias of Fourier feature networks: from regression to solving multi-scale PDEs with physics-informed neural networks
- Data-driven modeling of partially observed biological systems
- Gradient auxiliary physics-informed neural network for nonlinear biharmonic equation
- Extended dynamic mode decomposition for inhomogeneous problems
- Physics informed WNO
- PhyGeoNet: physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain
- Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
- Bayesian model and dimension reduction for uncertainty propagation: applications in random media
Uses Software
This page was built for publication: Simulator-free solution of high-dimensional stochastic elliptic partial differential equations using deep neural networks
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2223019)
