logo SBA

ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-03232021-104126


Tipo di tesi
Tesi di laurea magistrale
Autore
ANASTASI, GIADA
URN
etd-03232021-104126
Titolo
Design and implementation of Federated Clustering Algorithms
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
COMPUTER ENGINEERING
Relatori
relatore Prof. Marcelloni, Francesco
relatore Prof. Ducange, Pietro
relatore Ing. Renda, Alessandro
Parole chiave
  • clustering
  • dbscan
  • federated learning
  • fuzzy cmeans
  • kmeans
Data inizio appello
30/04/2021
Consultabilità
Non consultabile
Data di rilascio
30/04/2091
Riassunto
My thesis work is in the federated learning scenario, where different devices, within the same network, contain different information. Therefore, in this scenario the need arises to implement a clustering model that does not involve the sharing of private information, but at the same time achieves the same results of the centralized version, where the server is aware of all the data.
The purpose of my thesis is therefore to define and implement federated versions of some of the most popular clustering algorithms currently available: k-means, fuzzy c-means and dbscan. The versions I have defined have been tested on horizontally and vertically partitioned datasets in order to simulate a federated environment.
File