Solving for high-dimensional committor functions using artificial neural networks
From MaRDI portal
Fokker-Planck equations (35Q84) PDEs with randomness, stochastic partial differential equations (35R60) Stochastic ordinary differential equations (aspects of stochastic analysis) (60H10) Numerical solutions to stochastic differential and integral equations (65C30) Learning and adaptive systems in artificial intelligence (68T05)
Abstract: In this note we propose a method based on artificial neural network to study the transition between states governed by stochastic processes. In particular, we aim for numerical schemes for the committor function, the central object of transition path theory, which satisfies a high-dimensional Fokker-Planck equation. By working with the variational formulation of such partial differential equation and parameterizing the committor function in terms of a neural network, approximations can be obtained via optimizing the neural network weights using stochastic algorithms. The numerical examples show that moderate accuracy can be achieved for high-dimensional problems.
Recommendations
- Point cloud discretization of Fokker-Planck operators for committor functions
- Solving Fokker-Planck equation using deep learning
- Solving parametric PDE problems with artificial neural networks
- Neural parametric Fokker-Planck equation
- Deep neural network approximations for solutions of PDEs based on Monte Carlo algorithms
Cites work
- 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
- Diffusion maps
- Diffusion Maps, Reduction Coordinates, and Low Dimensional Representation of Stochastic Systems
- Point cloud discretization of Fokker-Planck operators for committor functions
- Reactive trajectories and the transition path process
- Reducing the Dimensionality of Data with Neural Networks
- Solving the quantum many-body problem with artificial neural networks
- The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems
- Towards a theory of transition paths
Cited in
(59)- \textit{hp}-VPINNs: variational physics-informed neural networks with domain decomposition
- Extending transition path theory: periodically driven and finite-time dynamics
- Committor functions via tensor networks
- Meta-learning pseudo-differential operators with deep neural networks
- On some neural network architectures that can represent viscosity solutions of certain high dimensional Hamilton-Jacobi partial differential equations
- On generalized residual network for deep learning of unknown dynamical systems
- Int-Deep: a deep learning initialized iterative method for nonlinear problems
- SelectNet: self-paced learning for high-dimensional partial differential equations
- MIM: a deep mixed residual method for solving high-order partial differential equations
- A semigroup method for high dimensional elliptic PDEs and eigenvalue problems based on neural networks
- Neural networks enforcing physical symmetries in nonlinear dynamical lattices: the case example of the Ablowitz-Ladik model
- Computing the invariant distribution of randomly perturbed dynamical systems using deep learning
- Solving eigenvalue PDEs of metastable diffusion processes using artificial neural networks
- Modeling the dynamics of PDE systems with physics-constrained deep auto-regressive networks
- Solving electrical impedance tomography with deep learning
- Deep multiscale model learning
- Integral representations of rank two false theta functions and their modularity properties
- Data-driven method to learn the most probable transition pathway and stochastic differential equation
- Active learning based sampling for high-dimensional nonlinear partial differential equations
- Neural network architectures using min-plus algebra for solving certain high-dimensional optimal control problems and Hamilton-Jacobi PDEs
- Computing committors in collective variables via Mahalanobis diffusion maps
- An overview on deep learning-based approximation methods for partial differential equations
- Solving traveltime tomography with deep learning
- Point cloud discretization of Fokker-Planck operators for committor functions
- Solving parametric PDE problems with artificial neural networks
- Algorithms for solving high dimensional PDEs: from nonlinear Monte Carlo to machine learning
- A Deep Learning Method for Elliptic Hemivariational Inequalities
- Neural parametric Fokker-Planck equation
- Coupling rare event algorithms with data-based learned committor functions using the analogue Markov chain
- Local and global perspectives on diffusion maps in the analysis of molecular systems
- Butterfly-net: optimal function representation based on convolutional neural networks
- Deep Nitsche Method: Deep Ritz Method with Essential Boundary Conditions
- SwitchNet: a neural Network Model for forward and inverse scattering problems
- Multilevel Fine-Tuning: Closing Generalization Gaps in Approximation of Solution Maps under a Limited Budget for Training Data
- Randomized neural network with Petrov-Galerkin methods for solving linear and nonlinear partial differential equations
- A priori generalization error analysis of two-layer neural networks for solving high dimensional Schrödinger eigenvalue problems
- Computing non-equilibrium trajectories by a deep learning approach
- Predicting rare events using neural networks and short-trajectory data
- Optimal Reaction Coordinates: Variational Characterization and Sparse Computation
- A deep learning method for computing eigenvalues of the fractional Schrödinger operator
- Adaptive Importance Sampling Based on Fault Tree Analysis for Piecewise Deterministic Markov Process
- A Reduced Order Schwarz Method for Nonlinear Multiscale Elliptic Equations Based on Two-Layer Neural Networks
- Solving inverse problems with deep learning
- A finite expression method for solving high-dimensional committor problems
- Generalization error estimates of a machine learning method for solving high-dimensional Schrödinger eigenvalue problems
- IG-PINNs: interface-gated physics-informed neural networks for solving elliptic interface problems
- Estimating committor functions via deep adaptive sampling on rare transition paths
- A deep learning method for computing committor functions with adaptive sampling
- Sharp error estimates for target measure diffusion maps with applications to the committor problem
- Large deviation minimizers for stochastic partial differential equations with degenerate noise
- Interpolating between BSDEs and PINNs: deep learning for elliptic and parabolic boundary value problems
- Approximation of functionals by neural network without curse of dimensionality
- Computing transition pathways for the study of rare events using deep reinforcement learning
- Diffusion methods for generating transition paths
- DeepONet-based operator learning for quantifying escape dynamics and most probable transition paths in stochastic dynamical systems
- Weak adversarial networks for high-dimensional partial differential equations
- Variational training of neural network approximations of solution maps for physical models
- Overcoming the curse of dimensionality for some Hamilton-Jacobi partial differential equations via neural network architectures
- Machine learning from a continuous viewpoint. I
This page was built for publication: Solving for high-dimensional committor functions using artificial neural networks
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2319851)
