kth.sePublications KTH
Operational message
There are currently operational disruptions. Troubleshooting is in progress.
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Optimisation Algorithms for Federated Learning
KTH, School of Electrical Engineering and Computer Science (EECS).
KTH, School of Electrical Engineering and Computer Science (EECS).
2025 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

A more recent development within machine learning is a technique called federated learning. Instead of training a model on a central server, a model is instead collaboratively trained on several devices. The data on each device is not shared with the server but instead kept locally on the devices, which increases data privacy and security for the users. Federated learning was introduced with the optimisation algorithm Federated Averaging (FedAvg), that uses a local stochastic gradient descent before communicating the updated weights to the server. However, FedAvg struggles in setups where the data distribution is unbalanced, and therefore other optimisation algorithms have been developed to address these issues. In this project, we aim to test different optimisation algorithms and study their accuracy and loss on different datasets. Specifically, the focus has been on heterogeneous data and fairness of the model performance. The optimisers were analysed and compared on different levels of data heterogeneity. Our analysis identifies the optimisation algorithm best suited for certain setups and concludes the overall best performing algorithm to be FedAvg-M. 

Abstract [sv]

En nyare utveckling inom maskininlärning är en teknik som kallas federated learning. Istället för att träna en modell på en central server, tränas modellen istället kollaborativt av flera enheter. Den data som finns på varje enhet delas inte med servern, utan stannar istället kvar på enheten, vilket ökar säkerheten och integriteten för användarna. Federated learning introducerades med optimeringsalgoritmen Federated Averaging (FedAvg) som använder en lokal stokastisk gradientnedstigning innan den kommunicerar de uppdaterade vikterna till servern. Dock har FedAvg svårigheter i situationer då datadistributionen är ojämn, och därför har andra optimeringsalgoritmer utvecklats för att tackla dessa problem. I detta projekt ämnar vi att testa olika optimeringsalgoritmer och studera deras noggrannhet och förlust på olika dataset. Speciellt har fokusen varit på heterogena data och hur rättvist modellerna presterar. Optimeringsalgoritmerna analyserades och jämfördes för olika nivåer av dataheterogenitet. Vår analys identifierar optimeringsalgoritmen som är best lämpad för olika situationer och drar slutsatsen att den algoritmen som generellt presterar bäst är FedAvg-M.

Place, publisher, year, edition, pages
2025. , p. 453-459
Series
TRITA-EECS-EX ; 2025:144
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
URN: urn:nbn:se:kth:diva-376166OAI: oai:DiVA.org:kth-376166DiVA, id: diva2:2034529
Supervisors
Examiners
Projects
Kandidatexamensarbete i Elektroteknik 2025, EECS, KTHAvailable from: 2026-02-02 Created: 2026-02-02

Open Access in DiVA

fulltext(80627 kB)20 downloads
File information
File name FULLTEXT01.pdfFile size 80627 kBChecksum SHA-512
35ce0a386dafe4649eb99cbe0efdfed651a3c9044e3339612422234d17a7e8ec21d4fd4aa201500c3c7a8f57194994b78b3e0cfbd5319ecd49f18a5d8a7ff775
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Electrical Engineering, Electronic Engineering, Information Engineering

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 4057 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf