Decentralized Asynchronous Non-convex Stochastic Optimization on Directed Graphs

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F23%3A00363763" target="_blank" >RIV/68407700:21230/23:00363763 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1109/TCNS.2023.3242043" target="_blank" >https://doi.org/10.1109/TCNS.2023.3242043</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/TCNS.2023.3242043" target="_blank" >10.1109/TCNS.2023.3242043</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Decentralized Asynchronous Non-convex Stochastic Optimization on Directed Graphs

Popis výsledku v původním jazyce

We consider a decentralized stochastic optimization problem over a network of agents, modeled as a directed graph: Agents aim to asynchronously minimize the average of their individual losses (possibly non-convex), each one having access only to a noisy estimate of the gradient of its own function. We propose an asynchronous distributed algorithm for such a class of problems. The algorithm combines stochastic gradients with tracking in an asynchronous push-sum framework and obtains a sublinear convergence rate, matching the rate of the centralized stochastic gradient descent applied to the nonconvex minimization. Our experiments on a non-convex image classification task using convolutional neural network validate the convergence of our proposed algorithm across different number of nodes and graph connectivity percentages.

Název v anglickém jazyce

Decentralized Asynchronous Non-convex Stochastic Optimization on Directed Graphs

Popis výsledku anglicky

We consider a decentralized stochastic optimization problem over a network of agents, modeled as a directed graph: Agents aim to asynchronously minimize the average of their individual losses (possibly non-convex), each one having access only to a noisy estimate of the gradient of its own function. We propose an asynchronous distributed algorithm for such a class of problems. The algorithm combines stochastic gradients with tracking in an asynchronous push-sum framework and obtains a sublinear convergence rate, matching the rate of the centralized stochastic gradient descent applied to the nonconvex minimization. Our experiments on a non-convex image classification task using convolutional neural network validate the convergence of our proposed algorithm across different number of nodes and graph connectivity percentages.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/EF16_019%2F0000765" target="_blank" >EF16_019/0000765: Výzkumné centrum informatiky</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2023

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

IEEE Transactions on Control of Network Systems

ISSN

2325-5870

e-ISSN

2372-2533

Svazek periodika

10

Číslo periodika v rámci svazku

4

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

9

Strana od-do

1796-1804

Kód UT WoS článku

001130821600020

EID výsledku v databázi Scopus

2-s2.0-85148444791