Concentration of Contractive Stochastic Approximation and Reinforcement Learning
Stochastic Systems · 2022-12 · 12 (4) · 411-430
DOI: 10.1287/stsy.2022.0097 ↗Kurz gefasst
Fachartikel in der Zeitschrift Stochastic Systems, 2022-12. Verfasst von Siddharth Chandak und 2 weiteren Autor:innen. Die Arbeit wurde bisher 18-mal zitiert (Crossref) und nennt 28 Literaturangaben.
Abstract (Originalsprache)
Using a martingale concentration inequality, concentration bounds “from time n 0 on” are derived for stochastic approximation algorithms with contractive maps and both martingale difference and Markov noises. These are applied to reinforcement learning algorithms, in particular to asynchronous Q-learning and TD(0). Funding: V. S. Borkar was supported in part by a S. S. Bhatnagar Fellowship from the Council of Scientific and Industrial Research, Government of India.
Zitieren
APA
Chandak, S., Borkar, V. S., & Dodhia, P. (2022). Concentration of Contractive Stochastic Approximation and Reinforcement Learning. Stochastic Systems, 12(4), 411-430. https://doi.org/10.1287/stsy.2022.0097
Vancouver
Chandak S, Borkar VS, Dodhia P. Concentration of Contractive Stochastic Approximation and Reinforcement Learning. Stochastic Systems. 2022;12(4):411-430. doi:10.1287/stsy.2022.0097
BibTeX
@article{Chandak2022,
author = {Chandak, Siddharth and Borkar, Vivek S. and Dodhia, Parth},
title = {Concentration of Contractive Stochastic Approximation and Reinforcement Learning},
journal = {Stochastic Systems},
year = {2022},
volume = {12},
number = {4},
pages = {411--430},
doi = {10.1287/stsy.2022.0097}
}
EUVIMED