Duality and Policy Evaluation in Distributionally Robust Bayesian Diffusion Control

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Blanchet, Jose, Cheng, Jiayi, Ling, Yuewei, Liu, Hao, Liu, Yang
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917235785728000
author Blanchet, Jose
Cheng, Jiayi
Ling, Yuewei
Liu, Hao
Liu, Yang
author_facet Blanchet, Jose
Cheng, Jiayi
Ling, Yuewei
Liu, Hao
Liu, Yang
contents We study diffusion control problems under parameter uncertainty. Controllers based on plug-in estimation can be brittle due to potential distribution shifts. Bayesian control with a prior on the parameters offers a formulation with beliefs about such shifts. However, as with any Bayesian model, the prior may be misspecified. To mitigate misspecification and reduce over-pessimism compared to classical robust control approaches (e.g. \citet{hansen2008robustness}), we propose a distributionally robust Bayesian control (DRBC) formulation in which an adversary perturbs the prior within a divergence neighborhood of a baseline prior. We develop a strong duality result that reduces the distributionally robust prior evaluation to a low-dimensional optimization and yields a practical simulation-based policy evaluation and learning procedure with structured policy parameterizations. We validate the efficiency of the algorithm on a synthetic linear-quadratic control example and real-data portfolio selection.
format Preprint
id arxiv_https___arxiv_org_abs_2506_19294
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Duality and Policy Evaluation in Distributionally Robust Bayesian Diffusion Control
Blanchet, Jose
Cheng, Jiayi
Ling, Yuewei
Liu, Hao
Liu, Yang
Optimization and Control
Probability
Portfolio Management
Machine Learning
We study diffusion control problems under parameter uncertainty. Controllers based on plug-in estimation can be brittle due to potential distribution shifts. Bayesian control with a prior on the parameters offers a formulation with beliefs about such shifts. However, as with any Bayesian model, the prior may be misspecified. To mitigate misspecification and reduce over-pessimism compared to classical robust control approaches (e.g. \citet{hansen2008robustness}), we propose a distributionally robust Bayesian control (DRBC) formulation in which an adversary perturbs the prior within a divergence neighborhood of a baseline prior. We develop a strong duality result that reduces the distributionally robust prior evaluation to a low-dimensional optimization and yields a practical simulation-based policy evaluation and learning procedure with structured policy parameterizations. We validate the efficiency of the algorithm on a synthetic linear-quadratic control example and real-data portfolio selection.
title Duality and Policy Evaluation in Distributionally Robust Bayesian Diffusion Control
topic Optimization and Control
Probability
Portfolio Management
Machine Learning
url https://arxiv.org/abs/2506.19294