Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 11 min ago
PreprintarXiv licencearXiv

Viscosity-Informed Generative Actor-Critic for High-Dimensional Stochastic Optimal Control

Abstract

We introduce a method for approximating viscosity solutions of stationary degenerate elliptic Hamilton--Jacobi--Bellman equations on bounded domains arising in stochastic exit-time control. Viscosity enforcement is formulated as a min--max problem over an envelope-generated test family parameterized by symmetric positive definite matrices. Under structural and asymptotic assumptions, any uniform limit point of the value function approximations satisfies the viscosity inequalities on the sampled test family. Numerical experiments show that the proposed method reduces empirical viscosity violations and improves robustness under perturbed dynamics.

Related

  1. Ergodicity for the Two-Dimensional Stochastic Navier-Stokes Equations with Shear-Thickening Viscosity
    Christopher Bunting
    arXivPreprintarXiv licenceFull textProbabilityarXiv
  2. Viscosity Solutions for Singular HJB Equations: BSDE Representations and Stochastic Control
    Dirk Becherer, Nicolas Perkowski, Yuchen Sun +1
    arXivPreprintarXiv licenceFull textMathematical physicsProbabilityarXiv
  3. Optimal Reinsurance-Dividend Strategy with Fixed Transaction Costs in a Regime-Switching Brownian Risk Model: A Viscosity Solution to the Impulse Control Problem
    Wenyuan Wang, Zuo Quan Xu, Kaixin Yan
    arXivPreprintarXiv licenceFull textOptimisation & controlarXiv
  4. Existence and Uniqueness of Viscosity Solutions for Stochastic Hamilton--Jacobi--Bellman Equations
    Jinniao Qiu
    arXivPreprintarXiv licenceFull textAnalysis & PDEsOptimisation & controlarXiv
  5. Singular viscosity-exponent limit for one-dimensional compressible Navier-Stokes equations with far field vacuum
    Xingyang Zhang
    arXivPreprintarXiv licenceFull textAnalysis & PDEsarXiv
BibTeX
@misc{golpashin2026viscosityinformed,
  title = {{Viscosity-Informed Generative Actor-Critic for High-Dimensional Stochastic Optimal Control}},
  author = {Alen E. Golpashin and Gokul Puthumanaillam and Melkior Ornik and Bruce A. Conway},
  howpublished = {arXiv},
  year = {2026},
  month = sep,
  doi = {10.48550/arxiv.2605.15668},
  eprint = {2605.15668},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.15668},
}
RIS
TY  - UNPB
TI  - Viscosity-Informed Generative Actor-Critic for High-Dimensional Stochastic Optimal Control
AU  - Alen E. Golpashin
AU  - Gokul Puthumanaillam
AU  - Melkior Ornik
AU  - Bruce A. Conway
PY  - 2026
DA  - 2026/09/18
PB  - arXiv
DO  - 10.48550/arxiv.2605.15668
UR  - https://arxiv.org/abs/2605.15668
AB  - We introduce a method for approximating viscosity solutions of stationary degenerate elliptic Hamilton--Jacobi--Bellman equations on bounded domains arising in stochastic exit-time control. Viscosity enforcement is formulated as a min--max problem over an envelope-generated test family parameterized by symmetric positive definite matrices. Under structural and asymptotic assumptions, any uniform limit point of the value function approximations satisfies the viscosity inequalities on the sampled test family. Numerical experiments show that the proposed method reduces empirical viscosity violations and improves robustness under perturbed dynamics.
ER  -