A Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs

This paper investigates a constrained distributed optimization problem enabled by differential privacy where the underlying network is time-changing with unbalanced digraphs. To solve such a problem, we first propose a differentially private online distributed algorithm by injecting adaptively adjus...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Rong Hu, Binru Zhang
Formato:	article
Lenguaje:	EN
Publicado:	Hindawi Limited 2021
Materias:	Mathematics QA1-939
Acceso en línea:	https://doaj.org/article/cc4f81876be3408eacb60c3a1f4a9268
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:cc4f81876be3408eacb60c3a1f4a9268
record_format	dspace
spelling	oai:doaj.org-article:cc4f81876be3408eacb60c3a1f4a92682021-11-08T02:35:21ZA Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs2314-478510.1155/2021/6115451https://doaj.org/article/cc4f81876be3408eacb60c3a1f4a92682021-01-01T00:00:00Zhttp://dx.doi.org/10.1155/2021/6115451https://doaj.org/toc/2314-4785This paper investigates a constrained distributed optimization problem enabled by differential privacy where the underlying network is time-changing with unbalanced digraphs. To solve such a problem, we first propose a differentially private online distributed algorithm by injecting adaptively adjustable Laplace noises. The proposed algorithm can not only protect the privacy of participants without compromising a trusted third party, but also be implemented on more general time-varying unbalanced digraphs. Under mild conditions, we then show that the proposed algorithm can achieve a sublinear expected bound of regret for general local convex objective functions. The result shows that there is a trade-off between the optimization accuracy and privacy level. Finally, numerical simulations are conducted to validate the efficiency of the proposed algorithm.Rong HuBinru ZhangHindawi LimitedarticleMathematicsQA1-939ENJournal of Mathematics, Vol 2021 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Mathematics QA1-939
spellingShingle	Mathematics QA1-939 Rong Hu Binru Zhang A Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs
description	This paper investigates a constrained distributed optimization problem enabled by differential privacy where the underlying network is time-changing with unbalanced digraphs. To solve such a problem, we first propose a differentially private online distributed algorithm by injecting adaptively adjustable Laplace noises. The proposed algorithm can not only protect the privacy of participants without compromising a trusted third party, but also be implemented on more general time-varying unbalanced digraphs. Under mild conditions, we then show that the proposed algorithm can achieve a sublinear expected bound of regret for general local convex objective functions. The result shows that there is a trade-off between the optimization accuracy and privacy level. Finally, numerical simulations are conducted to validate the efficiency of the proposed algorithm.
format	article
author	Rong Hu Binru Zhang
author_facet	Rong Hu Binru Zhang
author_sort	Rong Hu
title	A Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs
title_short	A Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs
title_full	A Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs
title_fullStr	A Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs
title_full_unstemmed	A Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs
title_sort	privacy-masking learning algorithm for online distributed optimization over time-varying unbalanced digraphs
publisher	Hindawi Limited
publishDate	2021
url	https://doaj.org/article/cc4f81876be3408eacb60c3a1f4a9268
work_keys_str_mv	AT ronghu aprivacymaskinglearningalgorithmforonlinedistributedoptimizationovertimevaryingunbalanceddigraphs AT binruzhang aprivacymaskinglearningalgorithmforonlinedistributedoptimizationovertimevaryingunbalanceddigraphs AT ronghu privacymaskinglearningalgorithmforonlinedistributedoptimizationovertimevaryingunbalanceddigraphs AT binruzhang privacymaskinglearningalgorithmforonlinedistributedoptimizationovertimevaryingunbalanceddigraphs
_version_	1718443229326606336

A Privacy-Masking Learning Algorithm for Online Distributed Optimization over Time-Varying Unbalanced Digraphs

Ejemplares similares