Deep BSDE Solver on Bounded Domains Part I: General Loss Rate

2025/08/19 by Würschmidt, Maximilian
#58J65 #65N15 #65N75 #FOS: Mathematics #G.1.8 #G.3 #Numerical Analysis (math.NA) #Probability (math.PR)

paper · doi:10.48550/arxiv.2508.14215

Abstract

We consider a ramification of the deep BSDE loss functional designed to apply for BSDEs on bounded domains, i.e. with random (unbounded) time horizons. We derive a general convergence rate of the loss functional; precisely for a class of (randomly) weighted modifications of the functional. The rate is expressed in terms of the underlying discrete-time stepsize and a universal approximation distance.

Citations

• A Discretization Scheme for BSDEs with Random Time Horizon
• A deep solver for backward stochastic Volterra integral equations
• A brief review of the Deep BSDE method for solving high-dimensional partial differential equations
• Convergence of a Deep BSDE solver with jumps
• Deep Picard Iteration for High-Dimensional Nonlinear PDEs
• DRM Revisited: A Complete Error Analysis
• Solving Poisson Equations using Neural Walk-on-Spheres
• Generalized convergence of the deep BSDE method: a step towards fully-coupled FBSDEs and applications in stochastic control
• Deep learning scheme for forward utilities using ergodic BSDEs
• Stochastic approach for elliptic problems in perforated domains
• Rectified deep neural networks overcome the curse of dimensionality when approximating solutions of McKean--Vlasov stochastic differential equations
• Mathematical Introduction to Deep Learning: Methods, Implementations, and Theory
• Deep neural networks overcome the curse of dimensionality in the numerical approximation of semilinear partial differential equations
• Interpolating between BSDEs and PINNs: deep learning for elliptic and parabolic boundary value problems
• Error Analysis of Deep Ritz Methods for Elliptic Equations
• A learning scheme by sparse grids and Picard approximations for semilinear parabolic PDEs
• Deep neural network approximation for high-dimensional elliptic PDEs with boundary conditions
• Overcoming the curse of dimensionality in the numerical approximation of high-dimensional semilinear elliptic partial differential equations
• Space-time error estimates for deep neural network approximations for differential equations
• Neural networks-based backward scheme for fully nonlinear PDEs
• Deep backward schemes for high-dimensional nonlinear PDEs
• Forward-Backward Stochastic Neural Networks: Deep Learning of High-dimensional Partial Differential Equations
• Second order backward SDE with random terminal time
• The Deep Ritz method: A deep learning-based numerical algorithm for solving variational problems
• First time to exit of a continuous Itô process: general moment estimates and L1-convergence rate for discrete time approximations

Related

• A Discretization Scheme for BSDEs with Random Time Horizon
• Donsker-Type Theorem for BSDEs: Rate of Convergence
• Generalized convergence of the deep BSDE method: a step towards fully-coupled FBSDEs and applications in stochastic control
• Asymptotic Expansion as Prior Knowledge in Deep Learning Method for high\n dimensional BSDEs
• Convergence of a Deep BSDE solver with jumps