Solving elliptic equations with Brownian motion: bias reduction and temporal difference learning

DOI10.1007/S11009-021-09871-9zbMATH Open1491.65015arXiv2008.00144OpenAlexW3046589003MaRDI QIDQ2157396FDOQ2157396

Yanyan Li

Publication date: 28 July 2022

Published in: Methodology and Computing in Applied Probability (Search for Journal in Brave)

Abstract: The Feynman-Kac formula provides a way to understand solutions to elliptic partial differential equations in terms of expectations of continuous time Markov processes. This connection allows for the creation of numerical schemes for solutions based on samples of these Markov processes which have advantages over traditional numerical methods in some cases. However, na"ive numerical implementations suffer from statistical bias and sampling error. We present methods to discretize the stochastic process appearing in the Feynman-Kac formula that reduce the bias of the numerical scheme. We also propose using temporal difference learning to assemble information from random samples in a way that is more efficient than the traditional Monte Carlo method.

Full work available at URL: https://arxiv.org/abs/2008.00144

Recommendations

• Solution Bounds for Elliptic Partial Differential Equations via Feynman-Kac Representation
• Feynman-Kac representation of fully nonlinear PDEs and applications
• Numerical algorithms for the forward and backward fractional Feynman-Kac equations
• Multilevel Monte Carlo for the Feynman-Kac formula for the Laplace equation
• On nonlinear Feynman-Kac formulas for viscosity solutions of semilinear parabolic partial differential equations

zbMATH Keywords

Monte CarloBrownian motionelliptic equationFeynman-Kac formulatemporal difference learning

Mathematics Subject Classification ID

Monte Carlo methods (65C05) Probabilistic methods, particle methods, etc. for boundary value problems involving PDEs (65N75) Numerical solutions to stochastic differential and integral equations (65C30)

Cites Work

• DGM: a deep learning algorithm for solving partial differential equations
• A continuity correction for discrete barrier options
• Stopped diffusion processes: boundary corrections and overshoot
• An Extension of MATLAB to Continuous Functions and Operators
• Title not available (Why is that?)
• FEYNMAN–KAC FORMULAS FOR BLACK–SCHOLES-TYPE OPERATORS
• Brownian Motion
• Title not available (Why is that?)
• A Feynman-Kac path-integral implementation for Poisson's equation using an \(h\)-conditioned Green's function
• Numerical solution of the Robin problem of Laplace equations with a Feynman-Kac formula and reflecting Brownian motions
• Solving Dirichlet problems numerically using the Feynman-Kac representation
• Title not available (Why is that?)
• Computing the principal eigenvalue of the Laplace operator by a stochastic method
• Simulator-free solution of high-dimensional stochastic elliptic partial differential equations using deep neural networks
• Solving high-dimensional partial differential equations using deep learning
• Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
• Exact Monte Carlo solution of elliptic partial differential equations
• Regional Monte Carlo solution of elliptic partial differential equations
• A Monte Carlo method for Poisson's equation
• Hidden physics models: machine learning of nonlinear partial differential equations
• 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
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• Deep Hidden Physics Models: Deep Learning of Nonlinear Partial Differential Equations
• Multilevel Monte Carlo for the Feynman-Kac formula for the Laplace equation
• Solving high-dimensional eigenvalue problems using deep neural networks: a diffusion Monte Carlo like approach
• A derivative-free method for solving elliptic partial differential equations with deep neural networks

Cited In (2)

• Connecting stochastic optimal control and reinforcement learning
• Learning elliptic partial differential equations with randomized linear algebra

This page was built for publication: Solving elliptic equations with Brownian motion: bias reduction and temporal difference learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2157396)