A polynomial proxy model approach to verifiable decentralized federated learning
Tan Li, Samuel Cheng, Tak Lam Chan, Haibo Hu · 2024
Read A polynomial proxy model approach to verifiable decentralized federated learning by Tan Li, Samuel Cheng, Tak Lam Chan, Haibo Hu (2024) 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/pmc-11579416.txt.
Description
Tan Li and co-authors secure shared model training without a central server, in work from 2024. In this setup, called decentralized federated learning, each site trains locally and shares only updates. Proving those updates were computed honestly, without leaking private data, is hard. Random dropout steps make the proof harder still.
Their framework, ProxyZKP, pairs each private model with a simpler proxy built from polynomials. Proofs of honest work run on the proxy, which suits the proof math well. Tests showed proof times a third to a half faster than known methods. Added noise also blunted attacks that try to rebuild private data.
Catalog record
| CREATOR | Tan Li, Samuel Cheng, Tak Lam Chan, Haibo Hu |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 141 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |