Regret Bounds for Episodic Risk-Sensitive Linear Quadratic Regulator

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xu, Wenhao, Gao, Xuefeng, He, Xuedong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910825231417344
author Xu, Wenhao
Gao, Xuefeng
He, Xuedong
author_facet Xu, Wenhao
Gao, Xuefeng
He, Xuedong
contents Risk-sensitive linear quadratic regulator is one of the most fundamental problems in risk-sensitive optimal control. In this paper, we study online adaptive control of risk-sensitive linear quadratic regulator in the finite horizon episodic setting. We propose a simple least-squares greedy algorithm and show that it achieves $\widetilde{\mathcal{O}}(\log N)$ regret under a specific identifiability assumption, where $N$ is the total number of episodes. If the identifiability assumption is not satisfied, we propose incorporating exploration noise into the least-squares-based algorithm, resulting in an algorithm with $\widetilde{\mathcal{O}}(\sqrt{N})$ regret. To our best knowledge, this is the first set of regret bounds for episodic risk-sensitive linear quadratic regulator. Our proof relies on perturbation analysis of less-standard Riccati equations for risk-sensitive linear quadratic control, and a delicate analysis of the loss in the risk-sensitive performance criterion due to applying the suboptimal controller in the online learning process.
format Preprint
id arxiv_https___arxiv_org_abs_2406_05366
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Regret Bounds for Episodic Risk-Sensitive Linear Quadratic Regulator
Xu, Wenhao
Gao, Xuefeng
He, Xuedong
Machine Learning
Optimization and Control
Risk-sensitive linear quadratic regulator is one of the most fundamental problems in risk-sensitive optimal control. In this paper, we study online adaptive control of risk-sensitive linear quadratic regulator in the finite horizon episodic setting. We propose a simple least-squares greedy algorithm and show that it achieves $\widetilde{\mathcal{O}}(\log N)$ regret under a specific identifiability assumption, where $N$ is the total number of episodes. If the identifiability assumption is not satisfied, we propose incorporating exploration noise into the least-squares-based algorithm, resulting in an algorithm with $\widetilde{\mathcal{O}}(\sqrt{N})$ regret. To our best knowledge, this is the first set of regret bounds for episodic risk-sensitive linear quadratic regulator. Our proof relies on perturbation analysis of less-standard Riccati equations for risk-sensitive linear quadratic control, and a delicate analysis of the loss in the risk-sensitive performance criterion due to applying the suboptimal controller in the online learning process.
title Regret Bounds for Episodic Risk-Sensitive Linear Quadratic Regulator
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2406.05366