Computing the invariant distribution of randomly perturbed dynamical systems using deep learning

DOI10.1007/S10915-022-01844-5zbMATH Open1490.35502arXiv2110.11538OpenAlexW3210372908MaRDI QIDQ2149015FDOQ2149015

Weiqing Ren, Qianxiao Li, Bo Lin

Publication date: 24 June 2022

Published in: Journal of Scientific Computing (Search for Journal in Brave)

Abstract: The invariant distribution, which is characterized by the stationary Fokker-Planck equation, is an important object in the study of randomly perturbed dynamical systems. Traditional numerical methods for computing the invariant distribution based on the Fokker-Planck equation, such as finite difference or finite element methods, are limited to low-dimensional systems due to the curse of dimensionality. In this work, we propose a deep learning based method to compute the generalized potential, i.e. the negative logarithm of the invariant distribution multiplied by the noise. The idea of the method is to learn a decomposition of the force field, as specified by the Fokker-Planck equation, from the trajectory data. The potential component of the decomposition gives the generalized potential. The method can deal with high-dimensional systems, possibly with partially known dynamics. Using the generalized potential also allows us to deal with systems at low temperatures, where the invariant distribution becomes singular around the metastable states. These advantages make it an efficient method to analyze invariant distributions for practical dynamical systems. The effectiveness of the proposed method is demonstrated by numerical examples.

Full work available at URL: https://arxiv.org/abs/2110.11538

zbMATH Keywords

deep learningFokker-Planck equationinvariant distributiongeneralized potential

Mathematics Subject Classification ID

Fokker-Planck equations (35Q84) Simulation of dynamical systems (37M05) Applications to the sciences (65Z05)

Cites Work

• DGM: a deep learning algorithm for solving partial differential equations
• Discovering governing equations from data by sparse identification of nonlinear dynamical systems
• Random Perturbations of Dynamical Systems
• Numerical solution of stochastic differential equations with jumps in finance
• Solving parametric PDE problems with artificial neural networks
• Solving high-dimensional partial differential equations using deep learning
• Numerical solution of non-linear Fokker-Planck equation using finite differences method and the cubic spline functions
• Efficient statistically accurate algorithms for the Fokker-Planck equation in large dimensions
• The Deep Ritz Method: a deep learning-based numerical algorithm for solving variational problems
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• Solving for high-dimensional committor functions using artificial neural networks
• Beating the curse of dimension with accurate statistics for the Fokker–Planck equation in complex turbulent systems
• Numerical Methods for Ordinary Differential Equations
• Weak adversarial networks for high-dimensional partial differential equations
• Fokker–Planck equations in the modeling of socio-economic phenomena
• Solving Fokker-Planck equation using deep learning
• Solving Inverse Stochastic Problems from Discrete Particle Observations Using the Fokker--Planck Equation and Physics-Informed Neural Networks
• Numerical simulations of Fokker-Plank equation by variational iteration method

Cited In (2)

• Computing high-dimensional invariant distributions from noisy data
• Generative Adversarial Network for Probabilistic Forecast of Random Dynamical Systems

Uses Software

• Adam
• DGM

Recommendations

• Computations in dynamical systems via random perturbations 👍 👎
• Computing Lyapunov functions using deep neural networks 👍 👎
• Estimating invariant probability densities for dynamical systems 👍 👎
• Physics-Informed Probabilistic Learning of Linear Embeddings of Nonlinear Dynamics with Guaranteed Stability 👍 👎
• Deep learning via dynamical systems: an approximation perspective 👍 👎
• Learning dynamical systems in noise using convolutional neural networks 👍 👎
• Learning stochastic dynamics with statistics-informed neural network 👍 👎
• A machine learning method for computing quasi-potential of stochastic dynamical systems 👍 👎

This page was built for publication: Computing the invariant distribution of randomly perturbed dynamical systems using deep learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2149015)