Convergence of kernel and neural-network methods for Hamilton--Jacobi--Bellman equations on unbounded domains

Abstract

We derive error bounds from empirical squared residuals for two non-monotone, mesh-free approximations of Hamilton--Jacobi--Bellman equations on unbounded spatial domains: collocation with Wendland kernels under a native space norm constraint, and physics-informed neural networks with smooth activations under a Sobolev-norm constraint. The main step is to recover the pointwise residual control needed for viscosity-solution error estimates from the empirical loss, while quantifying the discretization or sampling error. For the kernel method, we prove uniform convergence on compact sets under classical and native space regularity and derive an a posteriori quantitative error bound for the unique bounded continuous viscosity solution. For the neural-network method, we prove uniform convergence on compact sets in probability and give a quantitative high-probability error bound under classical Sobolev regularity. The bounds are governed by the fill distance for kernels and by the network and sample sizes for neural networks. The proofs combine empirical-to-population estimates, an interpolation inequality that turns an L² residual into a sup norm bound, and a controlled diffusion estimate for the error caused by truncating the domain. The regularity assumptions used to obtain convergence from approximation properties are additional to viscosity wellposedness. Numerical experiments in one and two space dimensions illustrate the methods and the estimates.

The Effect of Quadrature on the Convergence of Policy Iteration for Hamilton-Jacobi-Bellman Equations
Thomas Hall, Iain Smears, Endre Süli +1
arXiv23 Jun 2026PreprintCC BY 4.0Full textNumerical analysis & scientific computingarXiv
Convergence rates in the periodic homogenization of vanishing-viscous Hamilton-Jacobi equations
Kai Qin
arXiv25 Sep 2026PreprintCC BY-NC-ND 4.0Full textAnalysis & PDEsarXiv
Last-Iterate Convergence of Stochastic Gradient Methods for Unbounded Problems
Kevin Kurian Thomas Vaidyan, Ahmet Alacaoglu, Michael P. Friedlander
arXiv5 Nov 2026PreprintarXiv licenceFull textOptimisation & control StatisticsarXiv
C^(sigma+alpha) regularity for fractional Hamilton-Jacobi equations, and rates of convergence in periodic homogenization
Adina Ciomaga, Andrei Rodriguez-Paredes, Erwin Topp
arXiv12 Oct 2026PreprintarXiv licenceFull textAnalysis & PDEsarXiv
Solution of a system of Hamilton-Jacobi equations on a network
Emily R. Hammett, Maria Teresa Chiri, Kenneth D. Czuprynski +1
arXiv9 Oct 2026PreprintarXiv licenceFull textNumerical analysis & scientific computingarXiv

Cite

BibTeX RIS

BibTeX

@misc{nakano2026convergence,
  title = {{Convergence of kernel and neural-network methods for Hamilton--Jacobi--Bellman equations on unbounded domains}},
  author = {Yumiharu Nakano},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.09299},
  eprint = {2610.09299},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.09299},
}

RIS

TY  - UNPB
TI  - Convergence of kernel and neural-network methods for Hamilton--Jacobi--Bellman equations on unbounded domains
AU  - Yumiharu Nakano
PY  - 2026
DA  - 2026/10/07
PB  - arXiv
DO  - 10.48550/arxiv.2610.09299
UR  - https://arxiv.org/abs/2610.09299
AB  - We derive error bounds from empirical squared residuals for two non-monotone, mesh-free approximations of Hamilton--Jacobi--Bellman equations on unbounded spatial domains: collocation with Wendland kernels under a native space norm constraint, and physics-informed neural networks with smooth activations under a Sobolev-norm constraint. The main step is to recover the pointwise residual control needed for viscosity-solution error estimates from the empirical loss, while quantifying the discretization or sampling error. For the kernel method, we prove uniform convergence on compact sets under classical and native space regularity and derive an a posteriori quantitative error bound for the unique bounded continuous viscosity solution. For the neural-network method, we prove uniform convergence on compact sets in probability and give a quantitative high-probability error bound under classical Sobolev regularity. The bounds are governed by the fill distance for kernels and by the network and sample sizes for neural networks. The proofs combine empirical-to-population estimates, an interpolation inequality that turns an L^2 residual into a sup norm bound, and a controlled diffusion estimate for the error caused by truncating the domain. The regularity assumptions used to obtain convergence from approximation properties are additional to viscosity wellposedness. Numerical experiments in one and two space dimensions illustrate the methods and the estimates.
ER  -

Abstract

Related

Cite