prikaz prve stranice dokumenta Raspodijeljeno udruženo učenje za okoline Interneta stvari
No public access
master's thesis
Raspodijeljeno udruženo učenje za okoline Interneta stvari
Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing, 2022. urn:nbn:hr:168:422556
Bartaković, Marko
University of Zagreb
Faculty of Electrical Engineering and Computing

Institutional repository: FER Repository

Cite this document

Bartaković, M. (2022). Raspodijeljeno udruženo učenje za okoline Interneta stvari (Master's thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:422556

Bartaković, Marko. "Raspodijeljeno udruženo učenje za okoline Interneta stvari." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2022. https://urn.nsk.hr/urn:nbn:hr:168:422556

Bartaković, Marko. "Raspodijeljeno udruženo učenje za okoline Interneta stvari." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2022. https://urn.nsk.hr/urn:nbn:hr:168:422556

Bartaković, M. (2022). 'Raspodijeljeno udruženo učenje za okoline Interneta stvari', Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 25 September 2024, https://urn.nsk.hr/urn:nbn:hr:168:422556

Bartaković M. Raspodijeljeno udruženo učenje za okoline Interneta stvari [Master's thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2022 [cited 2024 September 25] Available at: https://urn.nsk.hr/urn:nbn:hr:168:422556

M. Bartaković, "Raspodijeljeno udruženo učenje za okoline Interneta stvari", Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2022. Available at: https://urn.nsk.hr/urn:nbn:hr:168:422556

Metadata
TitleRaspodijeljeno udruženo učenje za okoline Interneta stvari
Title (english)Federated Learning for Internet of Things Environments
AuthorMarko Bartaković
MentorIvana Podnar Žarko (mentor)
Committee memberIvana Podnar Žarko (predsjednik povjerenstva)
Committee memberMario Kušek (član povjerenstva)
Committee memberKrunoslav Tržec (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
Zagreb
Defense date and country2022-07-12, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
Raspodijeljeno udruženo učenje relativno je nova paradigma u treniranju modela strojnog učenja. Za treniranje modela koriste se uređaji na rubu mreže koji imaju vlastiti skup podataka za treniranje. Kako bi privatnost podataka korisnika ostala sačuvana, lokalni podaci klijenata nikada se ne prenose preko mreže. Proces treniranja modela izvodi se tako da klijenti na svojem lokalnom skupu podataka provode ažuriranje globalnog modela, a centralni čvor agregira ažuriranja modela od klijenata te kreira jedan globalni model. U ovom radu istraženo je područje raspodijeljenog udruženog učenja i primjeri korištenja takve vrste paradigme strojnog učenja. Kreirana su tri različita modela koji su predviđali prinos usjeva soje te su ispitane razlike u odnosu na modele koji su dobiveni na klasični način primjenom centraliziranog modela strojnog učenja.
Abstract (english)
Federated learning is an emerging paradigm used in the process of training of Machine Learning models. Edge devices are used to train machine learning models based on their local data sets. To preserve the privacy of user data, local data sets are never transmitted over the network. During the model training process, clients update a global model using their local datasets, and then the central server aggregates model updates from the clients to create one global model. In this thesis a review on federated learning and applications that have emerged using the paradigm of federated learning has been explored. Three different models were created to predict soybean crop yields and we examined the differences between three different models that are trained on a centralized dataset and models that are trained using the federated learning principles.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:168:422556
Study programmeTitle: Computing
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
Public note
Created on2023-01-13 14:29:57