Learning to Collaborate Over Graphs: A Selective Federated Multi-Task Learning Approach The work is funded by the European Union through the project 6G-INTENSE (G.A no. 101139266).
Ahmed Elbakary, Chaouki Ben Issaid, Mehdi Bennis · 2025
Read Learning to Collaborate Over Graphs: A Selective Federated Multi-Task Learning Approach The work is funded by the European Union through the project 6G-INTENSE (G.A no. 101139266). by Ahmed Elbakary, Chaouki Ben Issaid, Mehdi Bennis (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-2506-10102.txt.
Description
Ahmed Elbakary, Chaouki Ben Issaid, and Mehdi Bennis present a federated learning method that lets similar clients teach each other while keeping data local, in this 2025 arXiv paper. Each client holds its own data and its own task, so blind sharing can hurt as much as help. The method aims to pool lessons only where they fit.
Instead of sending whole models to a server, clients send a compact anchor vector that sums up what their data taught them, plus a small shared output layer. A graph links clients by likeness and is updated as their data drift. A community-finding step then groups near-twins, and help flows only inside each group. On two mixed datasets, the method beat leading rivals while using less computation and chatter, and it treated clients more fairly.
Catalog record
| CREATOR | Ahmed Elbakary, Chaouki Ben Issaid, Mehdi Bennis |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 49 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-10102