The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Cet article étudie un problème d'optimisation en ligne distribué et contraint sur des réseaux de communication fortement connectés, où une fonction de coût local de chaque agent varie dans le temps en raison de facteurs environnementaux. Nous proposons une méthode de sous-gradient projeté en ligne distribuée sur des réseaux dirigés déséquilibrés. La performance de la méthode proposée est évaluée par un regret qui est défini par l'erreur entre le coût cumulé dans le temps et le coût de la stratégie optimale a posteriori. Nous montrons qu’une limite de regret logarithmique peut être obtenue pour des fonctions de coût fortement convexes. Nous démontrons également la validité de la méthode proposée à travers un exemple numérique d'estimation distribuée sur un champ de diffusion.
Makoto YAMASHITA
Osaka University
Naoki HAYASHI
Osaka University
Takeshi HATANAKA
Tokyo Institute of Technology
Shigemasa TAKAI
Osaka University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Makoto YAMASHITA, Naoki HAYASHI, Takeshi HATANAKA, Shigemasa TAKAI, "Logarithmic Regret for Distributed Online Subgradient Method over Unbalanced Directed Networks" in IEICE TRANSACTIONS on Fundamentals,
vol. E104-A, no. 8, pp. 1019-1026, August 2021, doi: 10.1587/transfun.2020EAP1111.
Abstract: This paper investigates a constrained distributed online optimization problem over strongly connected communication networks, where a local cost function of each agent varies in time due to environmental factors. We propose a distributed online projected subgradient method over unbalanced directed networks. The performance of the proposed method is evaluated by a regret which is defined by the error between the cumulative cost over time and the cost of the optimal strategy in hindsight. We show that a logarithmic regret bound can be achieved for strongly convex cost functions. We also demonstrate the validity of the proposed method through a numerical example on distributed estimation over a diffusion field.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.2020EAP1111/_p
Copier
@ARTICLE{e104-a_8_1019,
author={Makoto YAMASHITA, Naoki HAYASHI, Takeshi HATANAKA, Shigemasa TAKAI, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Logarithmic Regret for Distributed Online Subgradient Method over Unbalanced Directed Networks},
year={2021},
volume={E104-A},
number={8},
pages={1019-1026},
abstract={This paper investigates a constrained distributed online optimization problem over strongly connected communication networks, where a local cost function of each agent varies in time due to environmental factors. We propose a distributed online projected subgradient method over unbalanced directed networks. The performance of the proposed method is evaluated by a regret which is defined by the error between the cumulative cost over time and the cost of the optimal strategy in hindsight. We show that a logarithmic regret bound can be achieved for strongly convex cost functions. We also demonstrate the validity of the proposed method through a numerical example on distributed estimation over a diffusion field.},
keywords={},
doi={10.1587/transfun.2020EAP1111},
ISSN={1745-1337},
month={August},}
Copier
TY - JOUR
TI - Logarithmic Regret for Distributed Online Subgradient Method over Unbalanced Directed Networks
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1019
EP - 1026
AU - Makoto YAMASHITA
AU - Naoki HAYASHI
AU - Takeshi HATANAKA
AU - Shigemasa TAKAI
PY - 2021
DO - 10.1587/transfun.2020EAP1111
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E104-A
IS - 8
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - August 2021
AB - This paper investigates a constrained distributed online optimization problem over strongly connected communication networks, where a local cost function of each agent varies in time due to environmental factors. We propose a distributed online projected subgradient method over unbalanced directed networks. The performance of the proposed method is evaluated by a regret which is defined by the error between the cumulative cost over time and the cost of the optimal strategy in hindsight. We show that a logarithmic regret bound can be achieved for strongly convex cost functions. We also demonstrate the validity of the proposed method through a numerical example on distributed estimation over a diffusion field.
ER -