Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Lee, Donghwan
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2605.16103
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917500515516416
author Lee, Donghwan
author_facet Lee, Donghwan
contents This paper develops a sign-separated finite-time error analysis for constant step-size Q-learning. Starting from the switching-system representation, the error is decomposed into its componentwise negative and positive parts. The negative part is dominated by a lower comparison linear time-invariant (LTI) system associated with a fixed optimal policy, whereas the positive part is controlled by a linear switching system. The resulting bounds show that the negative-side LTI certificate is no slower than the positive-side switching certificate and may produce a faster exponential envelope. The analysis identifies a max-induced asymmetry in Q-learning error dynamics. This asymmetry is connected to overestimation: positive action-wise errors can be selected and propagated by the Bellman maximum, whereas negative errors admit an optimal-policy lower comparison. Finite-time bounds are provided for both deterministic and stochastic constant-step-size recursions.
format Preprint
id arxiv_https___arxiv_org_abs_2605_16103
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Sign-Separated Finite-Time Error Analysis of Q-Learning
Lee, Donghwan
Artificial Intelligence
This paper develops a sign-separated finite-time error analysis for constant step-size Q-learning. Starting from the switching-system representation, the error is decomposed into its componentwise negative and positive parts. The negative part is dominated by a lower comparison linear time-invariant (LTI) system associated with a fixed optimal policy, whereas the positive part is controlled by a linear switching system. The resulting bounds show that the negative-side LTI certificate is no slower than the positive-side switching certificate and may produce a faster exponential envelope. The analysis identifies a max-induced asymmetry in Q-learning error dynamics. This asymmetry is connected to overestimation: positive action-wise errors can be selected and propagated by the Bellman maximum, whereas negative errors admit an optimal-policy lower comparison. Finite-time bounds are provided for both deterministic and stochastic constant-step-size recursions.
title Sign-Separated Finite-Time Error Analysis of Q-Learning
topic Artificial Intelligence
url https://arxiv.org/abs/2605.16103