prikaz prve stranice dokumenta Federalno učenje u uređajima na rubu temeljeno na grupama
No public access
undergraduate thesis
Federalno učenje u uređajima na rubu temeljeno na grupama
Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing, 2024. urn:nbn:hr:168:501831
Krklec, Andrija
University of Zagreb
Faculty of Electrical Engineering and Computing

Institutional repository: FER Repository

Cite this document

Krklec, A. (2024). Federalno učenje u uređajima na rubu temeljeno na grupama (Undergraduate thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:501831

Krklec, Andrija. "Federalno učenje u uređajima na rubu temeljeno na grupama." Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2024. https://urn.nsk.hr/urn:nbn:hr:168:501831

Krklec, Andrija. "Federalno učenje u uređajima na rubu temeljeno na grupama." Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2024. https://urn.nsk.hr/urn:nbn:hr:168:501831

Krklec, A. (2024). 'Federalno učenje u uređajima na rubu temeljeno na grupama', Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 26 December 2024, https://urn.nsk.hr/urn:nbn:hr:168:501831

Krklec A. Federalno učenje u uređajima na rubu temeljeno na grupama [Undergraduate thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2024 [cited 2024 December 26] Available at: https://urn.nsk.hr/urn:nbn:hr:168:501831

A. Krklec, "Federalno učenje u uređajima na rubu temeljeno na grupama", Undergraduate thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2024. Available at: https://urn.nsk.hr/urn:nbn:hr:168:501831

Metadata
TitleFederalno učenje u uređajima na rubu temeljeno na grupama
Title (english)Group-based federated learning on edge devices
AuthorAndrija Krklec
MentorIgor Čavrak (mentor)
Committee memberIgor Čavrak (član povjerenstva)
Committee memberDanko Basch (član povjerenstva)
Committee memberIvana Bosnić (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
Zagreb
Defense date and country2024-07-04, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
U ovom radu istražuju se osnovni koncepti federalnog učenja. Definira se pojam federalnog učenja i njegova podjela. Predstavlja se način federalnog učenja temeljen na grupama. Grupe se sastoje od klijenata s homogenim distribucijama podataka, nad kojima se zasebno provodi učenje. Ispituju se performanse različitih algoritama federalnog učenja FedAvg, FlexCFL, FeSEM i IFCA nad skupovima podataka MNIST, FEMNIST, Fashion MNIST. Skupovi su modificirani tako da simuliraju statističku heterogenost koja se pojavljuje u praktičnoj primjeni. Komentiraju se rezultati testiranja i daju ideje za budući nastavak rada.
Abstract (english)
In this paper, the basic concepts of federated learning are explored. The concept of federated learning and its classification are defined. A method of clustered federated learning is presented. The groups consist of clients with homogeneous data distributions, over which learning is conducted separately. The performance of various federated learning algorithms FedAvg, FlexCFL, FeSEM, and IFCA is examined on the datasets MNIST, FEMNIST, and Fashion MNIST. The datasets have been modified to simulate the statistical heterogeneity that occurs in practical applications. The test results are discussed, and ideas for future work are provided.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:168:501831
Study programmeTitle: Computing
Study programme type: university
Study level: undergraduate
Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
Public note
Created on2024-10-12 21:13:09