Physics-informed generative adversarial networks for stochastic differential equations
From MaRDI portal
(Redirected from Publication:5214836)
Abstract: We developed a new class of physics-informed generative adversarial networks (PI-GANs) to solve in a unified manner forward, inverse and mixed stochastic problems based on a limited number of scattered measurements. Unlike standard GANs relying only on data for training, here we encoded into the architecture of GANs the governing physical laws in the form of stochastic differential equations (SDEs) using automatic differentiation. In particular, we applied Wasserstein GANs with gradient penalty (WGAN-GP) for its enhanced stability compared to vanilla GANs. We first tested WGAN-GP in approximating Gaussian processes of different correlation lengths based on data realizations collected from simultaneous reads at sparsely placed sensors. We obtained good approximation of the generated stochastic processes to the target ones even for a mismatch between the input noise dimensionality and the effective dimensionality of the target stochastic processes. We also studied the overfitting issue for both the discriminator and generator, and we found that overfitting occurs also in the generator in addition to the discriminator as previously reported. Subsequently, we considered the solution of elliptic SDEs requiring approximations of three stochastic processes, namely the solution, the forcing, and the diffusion coefficient. We used three generators for the PI-GANs, two of them were feed forward deep neural networks (DNNs) while the other one was the neural network induced by the SDE. Depending on the data, we employed one or multiple feed forward DNNs as the discriminators in PI-GANs. Here, we have demonstrated the accuracy and effectiveness of PI-GANs in solving SDEs for up to 30 dimensions, but in principle, PI-GANs could tackle very high dimensional problems given more sensor data with low-polynomial growth in computational cost.
Recommendations
- Learning generative neural networks with physics knowledge
- Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Generative ensemble regression: Learning particle dynamics from observations of ensembles with physics-informed deep generative models
- Learning in Modal Space: Solving Time-Dependent Stochastic PDEs Using Physics-Informed Neural Networks
Cites work
- Bayesian deep convolutional encoder-decoder networks for surrogate modeling and uncertainty quantification
- Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
- DGM: a deep learning algorithm for solving partial differential equations
- Discovering governing equations from data by sparse identification of nonlinear dynamical systems
- Hidden physics models: machine learning of nonlinear partial differential equations
- scientific article; zbMATH DE number 6982909 (Why is no real title available?)
- Inferring solutions of differential equations using noisy multi-fidelity data
- Inverse problems: a Bayesian perspective
- Machine learning of linear differential equations using Gaussian processes
- Neural-net-induced Gaussian process regression for function approximation and PDE solution
- Numerical Gaussian processes for time-dependent and nonlinear partial differential equations
- Physics-informed cokriging: a Gaussian-process-regression-based multifidelity method for data-model convergence
- Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
- Solving parametric PDE problems with artificial neural networks
Cited in
(only showing first 100 items - show all)- A novel sequential method to train physics informed neural networks for Allen Cahn and Cahn Hilliard equations
- Solution of physics-based Bayesian inverse problems with deep generative priors
- Efficient coupled deep neural networks for the time-dependent coupled Stokes-Darcy problems
- Schwarz waveform relaxation-learning for advection-diffusion-reaction equations
- B-PINNs: Bayesian physics-informed neural networks for forward and inverse PDE problems with noisy data
- Physics-informed semantic inpainting: application to geostatistical modeling
- PhyGeoNet: physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain
- ISALT: inference-based schemes adaptive to large time-stepping for locally Lipschitz ergodic systems
- Adversarial sampling of unknown and high-dimensional conditional distributions
- Physics constrained learning for data-driven inverse modeling from sparse observations
- Learning functional priors and posteriors from data and physics
- Normalizing field flows: solving forward and inverse stochastic differential equations using physics-informed flow models
- CAN-PINN: a fast physics-informed neural network based on coupled-automatic-numerical differentiation method
- Learning generative neural networks with physics knowledge
- Error estimates for deep learning methods in fluid dynamics
- Surrogate and inverse modeling for two-phase flow in porous media via theory-guided convolutional neural network
- Scientific machine learning through physics-informed neural networks: where we are and what's next
- Data-driven rogue waves and parameters discovery in nearly integrable \(\mathcal{PT}\)-symmetric Gross-Pitaevskii equations via PINNs deep learning
- Deep neural networks based temporal-difference methods for high-dimensional parabolic partial differential equations
- Physics-informed neural networks for high-speed flows
- Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
- Conditional deep surrogate models for stochastic, high-dimensional, and multi-fidelity systems
- A-PINN: auxiliary physics informed neural networks for forward and inverse problems of nonlinear integro-differential equations
- Wasserstein generative adversarial uncertainty quantification in physics-informed neural networks
- Integrated finite element neural network (I-FENN) for non-local continuum damage mechanics
- PI-VAE: physics-informed variational auto-encoder for stochastic differential equations
- Uncertainty quantification in scientific machine learning: methods, metrics, and comparisons
- Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios
- Neural network architectures using min-plus algebra for solving certain high-dimensional optimal control problems and Hamilton-Jacobi PDEs
- Deep physics corrector: a physics enhanced deep learning architecture for solving stochastic differential equations
- An overview on deep learning-based approximation methods for partial differential equations
- Generative ensemble regression: Learning particle dynamics from observations of ensembles with physics-informed deep generative models
- Generative Adversarial Network for Probabilistic Forecast of Random Dynamical Systems
- Solving time dependent Fokker-Planck equations via temporal normalizing flow
- DeepXDE: a deep learning library for solving differential equations
- GAN-based priors for quantifying uncertainty in supervised learning
- SINDy-PI: a robust algorithm for parallel implicit sparse identification of nonlinear dynamics
- Bayesian differential programming for robust systems identification under uncertainty
- Enforcing imprecise constraints on generative adversarial networks for emulating physical systems
- Learning in Modal Space: Solving Time-Dependent Stochastic PDEs Using Physics-Informed Neural Networks
- Time-dependent Dirac equation with physics-informed neural networks: computation and properties
- On the order of derivation in the training of physics-informed neural networks: case studies for non-uniform beam structures
- Three ways to solve partial differential equations with neural networks — A review
- Monte Carlo simulation of SDEs using GANs
- Adaptive deep density approximation for fractional Fokker-Planck equations
- Pre-training strategy for solving evolution equations based on physics-informed neural networks
- Variable linear transformation improved physics-informed neural networks to solve thin-layer flow problems
- Physics-guided training of GAN to improve accuracy in airfoil design synthesis
- Pre-training physics-informed neural network with mixed sampling and its application in high-dimensional systems
- A Normalizing Field Flow Induced Two-Stage Stochastic Homogenization Method for Random Composite Materials
- Solution of physics-based inverse problems using conditional generative adversarial networks with full gradient penalty
- 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
- NeuralUQ: A Comprehensive Library for Uncertainty Quantification in Neural Differential Equations and Operators
- Bayesian Deep Learning Framework for Uncertainty Quantification in Stochastic Partial Differential Equations
- Physics-informed variational inference for uncertainty quantification of stochastic differential equations
- Bi-fidelity modeling of uncertain and partially unknown systems using DeepONets
- Posteriori error neural network: a recovery type posteriori error estimator based on neural network for diffusion problems
- Finite basis physics-informed neural networks (FBPINNs): a scalable domain decomposition approach for solving differential equations
- Resolution-independent generative models based on operator learning for physics-constrained Bayesian inverse problems
- Neural Control of Parametric Solutions for High-Dimensional Evolution PDEs
- Zero coordinate shift: whetted automatic differentiation for physics-informed operator learning
- CEENs: causality-enforced evolutional networks for solving time-dependent partial differential equations
- Learning scattering waves via coupling physics-informed neural networks and their convergence analysis
- Non-diffusive neural network method for hyperbolic conservation laws
- Bright-dark rogue wave transition in coupled ab system via the physics-informed neural networks method
- Tipping points of evolving epidemiological networks: machine learning-assisted, data-driven effective modeling
- Physics-specialized neural network with hard constraints for solving multi-material diffusion problems
- Connecting GANs, mean-field games, and optimal transport
- Solving partial differential equations by LS-SVM
- f-PICNN: a physics-informed convolutional neural network for partial differential equations with space-time domain
- Deep learning method for finding eigenpairs in Sturm-Liouville eigenvalue problems
- SDYN-GANs: adversarial learning methods for multistep generative models for general order stochastic dynamics
- Deep surrogate model for learning Green's function associated with linear reaction-diffusion operator
- Uncertainty quantification for noisy inputs-outputs in physics-informed neural networks and neural operators
- The Fourier spectral transformer for efficient and generalizable nonlinear PDEs
- Score-based physics-informed learning framework for stochastic dynamics
- IG-PINNs: interface-gated physics-informed neural networks for solving elliptic interface problems
- Physics informed neural networks for solving predator-prey models
- Forecasting of spatiotemporal nonlinear dynamic systems by physics-informed neural networks with ResNet blocks
- Reconstruction and prediction of Volterra integral equations driven by Gaussian noise
- Multi-task learning enhanced deep neural networks for solving delay differential equations
- Influence of noninstantaneous impulsive on null controllability of Atangana-Baleanu fractional stochastic differential equations
- Solving Lane-Emden-type eigenvalue problems with physics-informed neural networks
- Bounded KRnet and its applications to density estimation and approximation
- A PINN-driven game-theoretic framework in limited data photoacoustic tomography
- Learning WENO for entropy stable schemes to solve conservation laws
- A novel deep convolutional surrogate model with incomplete solve loss for parameterized steady-state diffusion problems
- Accelerating space-time methods using physics-informed neural networks
- Efficient data-driven polarization learning for attosecond science and nonperturbative nonlinear optics
- Data-driven fractional algebraic system solver
- IS-PINN: integrating mathematical principles to solve inhomogeneous wave equations via physics-informed neural networks
- Data-driven stochastic closure modeling via conditional diffusion model and neural operator
- GoRINNs: Godunov-Riemann informed neural networks for learning hyperbolic conservation laws
- Multi-head physics-informed neural networks for learning functional priors and uncertainty quantification
- Adaptive trajectories sampling for solving PDEs with deep learning methods
- Informed normalized gradient flow method for parameterized Schrödinger operators: a case study on photonic graphene
- On the use of exponential basis functions in the PINN architecture: an enhanced solution approach for the Laplace, Helmholtz, and elasto-static equations
- Generative adversarial physics-informed neural networks for solving forward and inverse problem with small labeled samples
- Approximation of solution operators for high-dimensional PDEs
Describes a project that uses
Uses Software
This page was built for publication: Physics-informed generative adversarial networks for stochastic differential equations
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5214836)
