A Policy Gradient Framework for Stochastic Optimal Control Problems with Global Convergence Guarantee

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhou, Mo, Lu, Jianfeng
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916688742580224
author Zhou, Mo
Lu, Jianfeng
author_facet Zhou, Mo
Lu, Jianfeng
contents We consider policy gradient methods for stochastic optimal control problem in continuous time. In particular, we analyze the gradient flow for the control, viewed as a continuous time limit of the policy gradient method. We prove the global convergence of the gradient flow and establish a convergence rate under some regularity assumptions. The main novelty in the analysis is the notion of local optimal control function, which is introduced to characterize the local optimality of the iterate.
format Preprint
id arxiv_https___arxiv_org_abs_2302_05816
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Policy Gradient Framework for Stochastic Optimal Control Problems with Global Convergence Guarantee
Zhou, Mo
Lu, Jianfeng
Optimization and Control
Machine Learning
Systems and Control
93E20 (Primary), 49L12 49M05 (secondary)
We consider policy gradient methods for stochastic optimal control problem in continuous time. In particular, we analyze the gradient flow for the control, viewed as a continuous time limit of the policy gradient method. We prove the global convergence of the gradient flow and establish a convergence rate under some regularity assumptions. The main novelty in the analysis is the notion of local optimal control function, which is introduced to characterize the local optimality of the iterate.
title A Policy Gradient Framework for Stochastic Optimal Control Problems with Global Convergence Guarantee
topic Optimization and Control
Machine Learning
Systems and Control
93E20 (Primary), 49L12 49M05 (secondary)
url https://arxiv.org/abs/2302.05816