Umeå universitets logga

umu.sePublikationer
Driftinformation
Ett driftavbrott i samband med versionsuppdatering är planerat till 17/11-2026, kl 12.00-13.00. Under den tidsperioden kommer DiVA inte att vara tillgängligt.
Ändra sökning
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • ieee
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
On the limits of privacy in federated learning: personalization and adversarial analysis
Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för datavetenskap. (NAUSICA group)ORCID-id: 0000-0003-0216-4992
2026 (Engelska)Doktorsavhandling, monografi (Övrigt vetenskapligt)
Abstract [en]

With the rapid adoption of Artificial Intelligence across diverse applications, concerns around data privacy have become increasingly prominent. Traditional centralized machine learning approaches require aggregating large volumes of user data at a central server, posing risks related to privacy, scalability, and regulatory compliance. Federated Learning has emerged as a decentralized alternative, enabling clients to train models locally while only sharing model updates with a central server. It addresses some of the concerns by enabling multiple participants to collaboratively train a shared model while keeping their raw data locally stored. However, data locality alone does not guarantee privacy. Information about local training records may still be revealed through model parameters, gradients, updates, or predictions.This makes FL vulnerable to privacy attacks such as model poisoning, data reconstruction and membership inference attacks. This thesis investigates these limitations in FL through two complementary perspectives:  adversarial evaluation of the framework and personalization of privacy protection. In this thesis, we first examine the validity of various data anonymization techniques for facial images through the re-identification attacks. A weight-optimized record linkage method is developed to determine whether anonymized image records can be linked to their corresponding original records. The evaluation considers different adversarial settings, including cases in which an attacker possesses only partial information. The findings demonstrate that visually modifying an image or separating identity-related information does not necessarily prevent re-identification. We also experiment with Shapley values to check whether they could be useful in detecting model poisoning attacks.Then, we develop two privacy-enhancing mechanisms for federated time-series forecasting. The first framework is centered around using differential privacy with transformers specifically for traffic flow data whereas the other framework is a microaggregation-based model agnostic k-anonymous federated forecasting framework. The results demonstrate that useful forecasting performance can be maintained while raw data remain decentralized and additional protection is applied at the client level. One of the major challenges that exist in FL is heterogeneity. There have been several studies in the literature on data and device heterogeneity in FL. However, the heterogeneity that exists in the privacy requirements of different clients is widely unexplored. Building on these findings, this thesis introduces a personalized privacy-preserving federated learning framework that allows individual clients to select their preferred privacy mechanism and protection level. The proposed FedPUP framework supports the simultaneous use of differential privacy and k-anonymity within a federated round. We also used Membership Inference attacks to empirically evaluate the information leakage associated with different privacy choices and to identify combinations of ε and k that provide comparable levels of protection. We also propose a privacy-aware server-side aggregation algorithm that incorporates client-level privacy characteristics during the aggregation process in each training round. Our empirical contributions are also complemented by an exploratory study investigating the practical requirements for deploying privacy-preserving federated learning systems in the industry. Our study highlights the importance of transparency, accountability, trust, costs, and privacy–utility trade-offs. 

Overall, this thesis demonstrates that federated learning should not be considered inherently private merely because data remain decentralized. By combining adversarial evaluation with personalized privacy-preserving mechanisms, it provides a more comprehensive understanding of privacy risks and safeguards in FL. The findings offer practical guidance for developing federated learning systems that are more privacy-aware, adaptable, and suitable for real-world deployment.

Ort, förlag, år, upplaga, sidor
Umeå: Umeå University, 2026. , s. 158
Serie
Report / UMINF, ISSN 0348-0542 ; 26.10
Nyckelord [en]
Federated Learning, Privacy, Differential Privacy, k-anonymity, Privacy attacks
Nationell ämneskategori
Datavetenskap (datalogi)
Forskningsämne
datalogi
Identifikatorer
URN: urn:nbn:se:umu:diva-259091ISBN: 978-91-6850-151-2 (tryckt)ISBN: 978-91-6850-152-9 (digital)OAI: oai:DiVA.org:umu-259091DiVA, id: diva2:2104299
Disputation
2026-10-16, NAT.D.470, Naturvetarhuset, 09:00 (Engelska)
Opponent
Handledare
Forskningsfinansiär
Wallenberg AI, Autonomous Systems and Software Program (WASP)Tillgänglig från: 2026-09-25 Skapad: 2026-09-23 Senast uppdaterad: 2026-09-24Bibliografiskt granskad

Open Access i DiVA

fulltext(9236 kB)393 nedladdningar
Filinformation
Filnamn FULLTEXT01.pdfFilstorlek 9236 kBChecksumma SHA-512
20aa17a371c4cfe8f4a1807abae7ae3ce83d7f33fa3b41df6713790edc773cf9b97f9951434ff9791f31eefb90b9acf58d1184ea4c15a75c4b81f4b2b1f6a04c
Typ fulltextMimetyp application/pdf
spikblad(202 kB)50 nedladdningar
Filinformation
Filnamn SPIKBLAD01.pdfFilstorlek 202 kBChecksumma SHA-512
343a3ad81bb9d033eab160d8497f81c68e98a6d059fd0a6be4f9dc119432eae759c1a8fd35bd3e3823168a5483de0a25257b5c77d0b8d44907d992b7ecfeb31a
Typ spikbladMimetyp application/pdf
cover(2408 kB)44 nedladdningar
Filinformation
Filnamn COVER01.pdfFilstorlek 2408 kBChecksumma SHA-512
9b9da1e3dd47b7b80803ad4359c1a8b8649361d7a0ac859771c9de2817be32866ad8580af3e651a31febf20c304e38a8659c4769575b59bcc12c3cb1d33c5e09
Typ coverMimetyp application/pdf

Person

Gupta, Sargam

Sök vidare i DiVA

Av författaren/redaktören
Gupta, Sargam
Av organisationen
Institutionen för datavetenskap
Datavetenskap (datalogi)

Sök vidare utanför DiVA

GoogleGoogle Scholar
Totalt: 393 nedladdningar
Antalet nedladdningar är summan av nedladdningar för alla fulltexter. Det kan inkludera t.ex tidigare versioner som nu inte längre är tillgängliga.

isbn
urn-nbn

Altmetricpoäng

isbn
urn-nbn
Totalt: 3388 träffar
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • ieee
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf