arXiv:2604.20492v1 Announce Type: new Abstract: In this paper, it is shown, for the first time, that centralized performance is achievable in decentralized learning without sharing the local datasets. Specifically, when clients adopt an empirical risk minimization with relative-entropy regularization (ERM-RER) learning framework and a forward-backward communication between clients is established,
Decentralized Machine Learning with Centralized Performance Guarantees via Gibbs Algorithms
Yaiza Bermudez, Samir Perlaza, I\~naki Esnaola·arXiv stat.ML··1 min read
a
Continue reading on arXiv stat.ML
This article was sourced from arXiv stat.ML's RSS feed. Visit the original for the complete story.