Non-Asymptotic Analysis of Online Local Private Learning with SGD Article Swipe

PDF

Enze Shi , Jinhan Xie , Bei Jiang , Linglong Kong , X. T. He ·

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2507.07041

Differentially Private Stochastic Gradient Descent (DP-SGD) has been widely used for solving optimization problems with privacy guarantees in machine learning and statistics. Despite this, a systematic non-asymptotic convergence analysis for DP-SGD, particularly in the context of online problems and local differential privacy (LDP) models, remains largely elusive. Existing non-asymptotic analyses have focused on non-private optimization methods, and hence are not applicable to privacy-preserving optimization problems. This work initiates the analysis to bridge this gap and opens the door to non-asymptotic convergence analysis of private optimization problems. A general framework is investigated for the online LDP model in stochastic optimization problems. We assume that sensitive information from individuals is collected sequentially and aim to estimate, in real-time, a static parameter that pertains to the population of interest. Most importantly, we conduct a comprehensive non-asymptotic convergence analysis of the proposed estimators in finite-sample situations, which gives their users practical guidelines regarding the effect of various hyperparameters, such as step size, parameter dimensions, and privacy budgets, on convergence rates. Our proposed estimators are validated in the theoretical and practical realms by rigorous mathematical derivations and carefully constructed numerical experiments.

Related Topics

Grand Theft Auto Online

How To Sell Drugs Online (Fast)

Multiplayer Online Battle Arena

Mathematical Analysis

Social Network Analysis

Massive Open Online Course

Complex Analysis

Technical Analysis

Concepts

No concepts available.

Metadata

Type: preprint
Language: en
Landing Page: http://arxiv.org/abs/2507.07041
PDF: https://arxiv.org/pdf/2507.07041
OA Status: green
OpenAlex ID: https://openalex.org/W4416063030

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4416063030

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2507.07041

Digital Object Identifier
Title: Non-Asymptotic Analysis of Online Local Private Learning with SGD

Work title
Type: preprint

OpenAlex work type
Language: en

Primary language
Publication year: 2025

Year of publication
Publication date: 2025-07-09

Full publication date if available
Authors: Enze Shi, Jinhan Xie, Bei Jiang, Linglong Kong, X. T. He

List of authors in order
Landing page: https://arxiv.org/abs/2507.07041

Publisher landing page
PDF URL: https://arxiv.org/pdf/2507.07041

Direct link to full text PDF
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2507.07041

Direct OA link when available
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W4416063030
doi	https://doi.org/10.48550/arxiv.2507.07041
ids.doi	https://doi.org/10.48550/arxiv.2507.07041
ids.openalex	https://openalex.org/W4416063030
fwci
type	preprint
title	Non-Asymptotic Analysis of Online Local Private Learning with SGD
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
is_xpac	False
apc_list
apc_paid
language	en
locations[0].id	pmh:oai:arXiv.org:2507.07041
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S4306400194
locations[0].source.issn
locations[0].source.type	repository
locations[0].source.is_oa	True
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	arXiv (Cornell University)
locations[0].source.host_organization	https://openalex.org/I205783295
locations[0].source.host_organization_name	Cornell University
locations[0].source.host_organization_lineage	https://openalex.org/I205783295
locations[0].license
locations[0].pdf_url	https://arxiv.org/pdf/2507.07041
locations[0].version	submittedVersion
locations[0].raw_type	text
locations[0].license_id
locations[0].is_accepted	False
locations[0].is_published	False
locations[0].raw_source_name
locations[0].landing_page_url	http://arxiv.org/abs/2507.07041
locations[1].id	doi:10.48550/arxiv.2507.07041
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published
locations[1].raw_source_name
locations[1].landing_page_url	https://doi.org/10.48550/arxiv.2507.07041
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5053922017
authorships[0].author.orcid
authorships[0].author.display_name	Enze Shi
authorships[0].author_position	first
authorships[0].raw_author_name	Shi, Enze
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5101055390
authorships[1].author.orcid	https://orcid.org/0000-0001-7407-8474
authorships[1].author.display_name	Jinhan Xie
authorships[1].author_position	middle
authorships[1].raw_author_name	Xie, Jinhan
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5006962593
authorships[2].author.orcid	https://orcid.org/0000-0002-0033-839X
authorships[2].author.display_name	Bei Jiang
authorships[2].author_position	middle
authorships[2].raw_author_name	Jiang, Bei
authorships[2].is_corresponding	False
authorships[3].author.id	https://openalex.org/A5062334200
authorships[3].author.orcid	https://orcid.org/0000-0003-3011-9216
authorships[3].author.display_name	Linglong Kong
authorships[3].author_position	middle
authorships[3].raw_author_name	Kong, Linglong
authorships[3].is_corresponding	False
authorships[4].author.id	https://openalex.org/A5100432159
authorships[4].author.orcid	https://orcid.org/0000-0002-7033-7640
authorships[4].author.display_name	X. T. He
authorships[4].author_position	last
authorships[4].raw_author_name	He, Xuming
authorships[4].is_corresponding	False
has_content.pdf	False
has_content.grobid_xml	False
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2507.07041
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-10T00:00:00
display_name	Non-Asymptotic Analysis of Online Local Private Learning with SGD
has_fulltext	False
is_retracted	False
updated_date	2025-11-28T10:24:58.072318
primary_topic
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2507.07041
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license
best_oa_location.pdf_url	https://arxiv.org/pdf/2507.07041
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2507.07041
primary_location.id	pmh:oai:arXiv.org:2507.07041
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S4306400194
primary_location.source.issn
primary_location.source.type	repository
primary_location.source.is_oa	True
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	arXiv (Cornell University)
primary_location.source.host_organization	https://openalex.org/I205783295
primary_location.source.host_organization_name	Cornell University
primary_location.source.host_organization_lineage	https://openalex.org/I205783295
primary_location.license
primary_location.pdf_url	https://arxiv.org/pdf/2507.07041
primary_location.version	submittedVersion
primary_location.raw_type	text
primary_location.license_id
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	http://arxiv.org/abs/2507.07041
publication_date	2025-07-09
publication_year	2025
referenced_works_count	0
abstract_inverted_index.A	86
abstract_inverted_index.a	24, 116, 130
abstract_inverted_index.We	100
abstract_inverted_index.as	155
abstract_inverted_index.by	177
abstract_inverted_index.in	17, 32, 96, 114, 139, 171
abstract_inverted_index.is	89, 107
abstract_inverted_index.of	35, 82, 124, 135, 151
abstract_inverted_index.on	52, 163
abstract_inverted_index.to	61, 70, 78, 112, 121
abstract_inverted_index.we	128
abstract_inverted_index.LDP	94
abstract_inverted_index.Our	166
abstract_inverted_index.aim	111
abstract_inverted_index.and	20, 38, 56, 74, 110, 160, 174, 181
abstract_inverted_index.are	58, 169
abstract_inverted_index.for	10, 29, 91
abstract_inverted_index.gap	73
abstract_inverted_index.has	6
abstract_inverted_index.not	59
abstract_inverted_index.the	33, 68, 76, 92, 122, 136, 149, 172
abstract_inverted_index.Most	126
abstract_inverted_index.This	65
abstract_inverted_index.been	7
abstract_inverted_index.door	77
abstract_inverted_index.from	105
abstract_inverted_index.have	50
abstract_inverted_index.step	156
abstract_inverted_index.such	154
abstract_inverted_index.that	102, 119
abstract_inverted_index.this	72
abstract_inverted_index.used	9
abstract_inverted_index.with	14
abstract_inverted_index.work	66
abstract_inverted_index.(LDP)	42
abstract_inverted_index.gives	143
abstract_inverted_index.hence	57
abstract_inverted_index.local	39
abstract_inverted_index.model	95
abstract_inverted_index.opens	75
abstract_inverted_index.size,	157
abstract_inverted_index.their	144
abstract_inverted_index.this,	23
abstract_inverted_index.users	145
abstract_inverted_index.which	142
abstract_inverted_index.assume	101
abstract_inverted_index.bridge	71
abstract_inverted_index.effect	150
abstract_inverted_index.online	36, 93
abstract_inverted_index.rates.	165
abstract_inverted_index.realms	176
abstract_inverted_index.static	117
abstract_inverted_index.widely	8
abstract_inverted_index.DP-SGD,	30
abstract_inverted_index.Descent	4
abstract_inverted_index.Despite	22
abstract_inverted_index.Private	1
abstract_inverted_index.conduct	129
abstract_inverted_index.context	34
abstract_inverted_index.focused	51
abstract_inverted_index.general	87
abstract_inverted_index.largely	45
abstract_inverted_index.machine	18
abstract_inverted_index.models,	43
abstract_inverted_index.privacy	15, 41, 161
abstract_inverted_index.private	83
abstract_inverted_index.remains	44
abstract_inverted_index.solving	11
abstract_inverted_index.various	152
abstract_inverted_index.(DP-SGD)	5
abstract_inverted_index.Existing	47
abstract_inverted_index.Gradient	3
abstract_inverted_index.analyses	49
abstract_inverted_index.analysis	28, 69, 81, 134
abstract_inverted_index.budgets,	162
abstract_inverted_index.elusive.	46
abstract_inverted_index.learning	19
abstract_inverted_index.methods,	55
abstract_inverted_index.pertains	120
abstract_inverted_index.problems	13, 37
abstract_inverted_index.proposed	137, 167
abstract_inverted_index.rigorous	178
abstract_inverted_index.carefully	182
abstract_inverted_index.collected	108
abstract_inverted_index.estimate,	113
abstract_inverted_index.framework	88
abstract_inverted_index.initiates	67
abstract_inverted_index.interest.	125
abstract_inverted_index.numerical	184
abstract_inverted_index.parameter	118, 158
abstract_inverted_index.practical	146, 175
abstract_inverted_index.problems.	64, 85, 99
abstract_inverted_index.regarding	148
abstract_inverted_index.sensitive	103
abstract_inverted_index.validated	170
abstract_inverted_index.Stochastic	2
abstract_inverted_index.applicable	60
abstract_inverted_index.estimators	138, 168
abstract_inverted_index.guarantees	16
abstract_inverted_index.guidelines	147
abstract_inverted_index.population	123
abstract_inverted_index.real-time,	115
abstract_inverted_index.stochastic	97
abstract_inverted_index.systematic	25
abstract_inverted_index.constructed	183
abstract_inverted_index.convergence	27, 80, 133, 164
abstract_inverted_index.derivations	180
abstract_inverted_index.dimensions,	159
abstract_inverted_index.individuals	106
abstract_inverted_index.information	104
abstract_inverted_index.non-private	53
abstract_inverted_index.situations,	141
abstract_inverted_index.statistics.	21
abstract_inverted_index.theoretical	173
abstract_inverted_index.differential	40
abstract_inverted_index.experiments.	185
abstract_inverted_index.importantly,	127
abstract_inverted_index.investigated	90
abstract_inverted_index.mathematical	179
abstract_inverted_index.optimization	12, 54, 63, 84, 98
abstract_inverted_index.particularly	31
abstract_inverted_index.sequentially	109
abstract_inverted_index.comprehensive	131
abstract_inverted_index.finite-sample	140
abstract_inverted_index.Differentially	0
abstract_inverted_index.non-asymptotic	26, 48, 79, 132
abstract_inverted_index.hyperparameters,	153
abstract_inverted_index.privacy-preserving	62
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	5
citation_normalized_percentile