Diversity-Driven Learning: Tackling Spurious Correlations and Data Heterogeneity in Federated Models

ely D. Németh [email protected], Fanì, Jeng Ng, ara Caputo, el Ángel Lozano, a Oliver, Quadrianto · 2025

Read Diversity-Driven Learning: Tackling Spurious Correlations and Data Heterogeneity in Federated Models by ely D. Németh [email protected], Fanì, Jeng Ng, ara Caputo, el Ángel Lozano, a Oliver, Quadrianto (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-2504-11216.txt.

Description

Gergely Németh and co-authors tackle a weak point in 2025 in federated learning, where many devices train one shared model without sending their raw data to a center. Each device holds data that differs from the rest, in mix and in balance. That spread slows training and hurts how well the shared model works for all.

Their method, called FedDiverse, picks which clients take part in each round based on how varied their data is. First, the team sets out six measures that capture the kinds of imbalance at play, from class mix to false links in the data. Then they built seven image data sets to test client choice under those strains. Picking a varied set of clients for each round aims to keep the shared model fair and strong across all devices.

Catalog record

CREATORely D. Németh [email protected], Fanì, Jeng Ng, ara Caputo, el Ángel Lozano, a Oliver, Quadrianto
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE66 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-11216

Browse the full library · About the Tower of Babel