Interaction-Aware Gaussian Weighting for Clustered Federated Learning

Alessandro Licciardi, Davide Leo, Eros Faní, Barbara Caputo, Marco Ciccone · 2025

Read Interaction-Aware Gaussian Weighting for Clustered Federated Learning by Alessandro Licciardi, Davide Leo, Eros Faní, Barbara Caputo, Marco Ciccone (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-03340.txt.

Description

Alessandro Licciardi and four co-authors improve clustered federated learning, a way to train models across many devices without gathering their data. When users' data differ wildly, one shared model fits nobody well. Clustering answers this by grouping similar clients and training a model per group. The hard part is judging similarity without seeing the data.

Their method, FedGWC, turns each client's training loss into a signal about which other clients it resembles, using a Gaussian weighting scheme to reward close matches. They also propose a new score that rates how tight a cluster is, taking class balance into account through a distance measure between distributions. On standard benchmark datasets the method formed better clusters and classified more accurately than existing approaches. Privacy stays intact throughout.

Catalog record

CREATORAlessandro Licciardi, Davide Leo, Eros Faní, Barbara Caputo, Marco Ciccone
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE77 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-03340

Browse the full library · About the Tower of Babel