Solving high-dimensional partial differential equations using deep learning

DOI10.1073/PNAS.1718942115zbMATH Open1416.35137arXiv1707.02568OpenAlexW2803629276WikidataQ64123028 ScholiaQ64123028MaRDI QIDQ4967451FDOQ4967451

Authors: Jiequn Han, Arnulf Jentzen, Weinan E 

Publication date: 3 July 2019

Published in: Proceedings of the National Academy of Sciences (Search for Journal in Brave)

Abstract: Developing algorithms for solving high-dimensional partial differential equations (PDEs) has been an exceedingly difficult task for a long time, due to the notoriously difficult problem known as the "curse of dimensionality". This paper introduces a deep learning-based approach that can handle general high-dimensional parabolic PDEs. To this end, the PDEs are reformulated using backward stochastic differential equations and the gradient of the unknown solution is approximated by neural networks, very much in the spirit of deep reinforcement learning with the gradient acting as the policy function. Numerical results on examples including the nonlinear Black-Scholes equation, the Hamilton-Jacobi-Bellman equation, and the Allen-Cahn equation suggest that the proposed algorithm is quite effective in high dimensions, in terms of both accuracy and cost. This opens up new possibilities in economics, finance, operational research, and physics, by considering all participating agents, assets, resources, or particles together at the same time, instead of making ad hoc assumptions on their inter-relationships.

Full work available at URL: https://arxiv.org/abs/1707.02568

Recommendations

• Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
• Solving Kolmogorov PDEs without the curse of dimensionality via deep learning and asymptotic expansion with Malliavin calculus
• Solving non-linear Kolmogorov equations in large dimensions by using deep learning: a numerical comparison of discretization schemes
• Algorithms for solving high dimensional PDEs: from nonlinear Monte Carlo to machine learning
• DGM: a deep learning algorithm for solving partial differential equations

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Neural biology (92C20) Nonlinear parabolic equations (35K55) Rationality and learning in game theory (91A26)

Cited In (only showing first 100 items - show all)

• Low-dimensional approximations of high-dimensional asset price models
• Lookback option pricing under the double Heston model using a deep learning algorithm
• Committor functions via tensor networks
• Deep learning for constrained utility maximisation
• Solving parametric partial differential equations with deep rectified quadratic unit neural networks
• The backward Euler-Maruyama method for invariant measures of stochastic differential equations with super-linear coefficients
• A physics-informed learning approach to Bernoulli-type free boundary problems
• A numerical approach to Kolmogorov equation in high dimension based on Gaussian analysis
• DNN expression rate analysis of high-dimensional PDEs: application to option pricing
• A discontinuity capturing shallow neural network for elliptic interface problems
• DeepM\&Mnet: inferring the electroconvection multiphysics fields based on operator approximation by neural networks
• Solving inverse stochastic problems from discrete particle observations using the Fokker-Planck equation and physics-informed neural networks
• Solving time dependent Fokker-Planck equations via temporal normalizing flow
• PFNN-2: a domain decomposed penalty-free neural network method for solving partial differential equations
• Numerical solution and bifurcation analysis of nonlinear partial differential equations with extreme learning machines
• A new deep neural network algorithm for multiple stopping with applications in options pricing
• On stability and regularization for data-driven solution of parabolic inverse source problems
• Structure preservation for the deep neural network multigrid solver
• Deep learning of free boundary and Stefan problems
• MIM: a deep mixed residual method for solving high-order partial differential equations
• Computing the invariant distribution of randomly perturbed dynamical systems using deep learning
• Fractional Chebyshev deep neural network (FCDNN) for solving differential models
• A nonlocal physics-informed deep learning framework using the peridynamic differential operator
• An efficient weak Euler-Maruyama type approximation scheme of very high dimensional SDEs by orthogonal random variables
• Kolmogorov n-width and Lagrangian physics-informed neural networks: a causality-conforming manifold for convection-dominated PDEs
• Meta-learning pseudo-differential operators with deep neural networks
• Using deep learning to extend the range of air pollution monitoring and forecasting
• A probabilistic generative model for semi-supervised training of coarse-grained surrogates and enforcing physical constraints through virtual observables
• Structure probing neural network deflation
• On generalized residual network for deep learning of unknown dynamical systems
• A derivative-free method for solving elliptic partial differential equations with deep neural networks
• Learning finite element convergence with the multi-fidelity graph neural network
• Enforcing constraints for interpolation and extrapolation in generative adversarial networks
• An overview on deep learning-based approximation methods for partial differential equations
• ISALT: inference-based schemes adaptive to large time-stepping for locally Lipschitz ergodic systems
• An adaptive sparse grid local discontinuous Galerkin method for Hamilton-Jacobi equations in high dimensions
• SelectNet: self-paced learning for high-dimensional partial differential equations
• A physics-informed convolutional neural network for the simulation and prediction of two-phase Darcy flows in heterogeneous porous media
• Approximation properties of residual neural networks for Kolmogorov PDEs
• An unsupervised deep learning approach to solving partial integro-differential equations
• Accelerated share repurchase and other buyback programs: what neural networks can bring
• Solving Allen-Cahn and Cahn-Hilliard Equations using the Adaptive Physics Informed Neural Networks
• Physics-Informed Neural Networks for Solving Dynamic Two-Phase Interface Problems
• Algorithms for solving high dimensional PDEs: from nonlinear Monte Carlo to machine learning
• Learning and meta-learning of stochastic advection–diffusion–reaction systems from sparse measurements
• Multi-Scale Deep Neural Network (MscaleDNN) Methods for Oscillatory Stokes Flows in Complex Domains
• A Local Deep Learning Method for Solving High Order Partial Differential Equations
• Sparse Deep Neural Network for Nonlinear Partial Differential Equations
• Approximations with deep neural networks in Sobolev time-space
• Deep xVA Solver: A Neural Network–Based Counterparty Credit Risk Management Framework
• Galerkin neural networks: a framework for approximating variational equations with error control
• Multi-scale deep neural network (MscaleDNN) for solving Poisson-Boltzmann equation in complex domains
• An adaptive surrogate modeling based on deep neural networks for large-scale Bayesian inverse problems
• Physics informed neural networks (PINNs) for approximating nonlinear dispersive PDEs
• Deep splitting method for parabolic PDEs
• Neural network approach to data-driven estimation of chemotactic sensitivity in the Keller-Segel model
• An application of the splitting-up method for the computation of a neural network representation for the solution for the filtering equations
• Deep combinatorial optimisation for optimal stopping time problems: application to swing options pricing.
• Efficient coupled deep neural networks for the time-dependent coupled Stokes-Darcy problems
• Neural networks enforcing physical symmetries in nonlinear dynamical lattices: the case example of the Ablowitz-Ladik model
• Extensions of the deep Galerkin method
• Unbiased deep solvers for linear parametric PDEs
• BCR-net: A neural network based on the nonstandard wavelet form
• Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data
• Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
• Solving high-dimensional Hamilton-Jacobi-Bellman PDEs using neural networks: perspectives from the theory of controlled diffusions and measures on path space
• Optimizing a portfolio of mean-reverting assets with transaction costs via a feedforward neural network
• Extreme learning machine collocation for the numerical solution of elliptic PDEs with sharp gradients
• Deep autoencoder based energy method for the bending, vibration, and buckling analysis of Kirchhoff plates with transfer learning
• Deep learning methods for partial differential equations and related parameter identification problems
• Approximation rates for neural networks with encodable weights in smoothness spaces
• A deep learning approach to the probabilistic numerical solution of path-dependent partial differential equations
• On a multilevel Levenberg-Marquardt method for the training of artificial neural networks and its application to the solution of partial differential equations
• Learning constitutive relations using symmetric positive definite neural networks
• Deep Neural Networks for Solving Large Linear Systems Arising from High-Dimensional Problems
• Tensor decomposition methods for high-dimensional Hamilton-Jacobi-Bellman equations
• Machine learning from a continuous viewpoint. I
• A proof that rectified deep neural networks overcome the curse of dimensionality in the numerical approximation of semilinear heat equations
• Scalable uncertainty quantification for deep operator networks using randomized priors
• Learning nonlocal constitutive models with neural networks
• Learning viscoelasticity models from indirect data using deep neural networks
• Train Like a (Var)Pro: Efficient Training of Neural Networks with Variable Projection
• Overcoming the curse of dimensionality in the numerical approximation of parabolic partial differential equations with gradient-dependent nonlinearities
• Machine learning for semi linear PDEs
• Data-driven hedging of stock index options via deep learning
• Machine learning and computational mathematics
• Numerical simulation of Volterra-Fredholm integral equations using least squares support vector regression
• Deep unfitted Nitsche method for elliptic interface problems
• A fully nonlinear Feynman-Kac formula with derivatives of arbitrary orders
• Numerical solution of a class of nonlinear partial differential equations by using barycentric interpolation collocation method
• A survey of numerical solutions for stochastic control problems: some recent progress
• Numerical solution of the Fokker-Planck equation using physics-based mixture models
• DeepXDE: a deep learning library for solving differential equations
• Adaptive deep learning for high-dimensional Hamilton-Jacobi-Bellman equations
• Path-dependent deep Galerkin method: a neural network approach to solve path-dependent partial differential equations
• Actor-critic method for high dimensional static Hamilton-Jacobi-Bellman partial differential equations based on neural networks
• Analysis of the generalization error: empirical risk minimization over deep artificial neural networks overcomes the curse of dimensionality in the numerical approximation of Black-Scholes partial differential equations
• State-dependent Riccati equation feedback stabilization for nonlinear PDEs
• Trend to equilibrium for the kinetic Fokker-Planck equation via the neural network approach
• Overcoming the curse of dimensionality in the numerical approximation of Allen-Cahn partial differential equations via truncated full-history recursive multilevel Picard approximations

Uses Software

• TensorFlow

This page was built for publication: Solving high-dimensional partial differential equations using deep learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4967451)