Asynchronous Gossip Algorithms for Rank-Based Statistical Methods

Anna van Elst, Igor Colin, Stephan Clémençon · 2025

Read Asynchronous Gossip Algorithms for Rank-Based Statistical Methods by Anna van Elst, Igor Colin, Stephan Clémençon (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-2509-07543.txt.

Description

Anna Van Elst and 2 co-authors develop gossip algorithms in 2025 for computing robust statistics across a decentralized network. Gossip algorithms spread information by repeated local exchanges between neighbors, with no central server. Decentralized AI and edge systems need them, but classic versions rely on means and sums that corrupted or adversarial data can easily poison.

The new algorithms compute a broad class of rank-based statistics, including L-statistics and rank statistics, which resist outliers because they depend on ordering rather than size. The team applies them to robust distributed two-sample testing and gives the first gossip algorithm for Wilcoxon rank-sum tests. They prove convergence guarantees, including the first rate bound for asynchronous gossip-based rank estimation. Experiments on diverse network topologies confirm the theory, supporting trustworthy statistics on edge networks.

Catalog record

CREATORAnna van Elst, Igor Colin, Stephan Clémençon
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE47 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-07543

Browse the full library · About the Tower of Babel