Improved Stability and Generalization Guarantees of the Decentralized SGD Algorithm
Batiste Le bars, Aurélien Bellet, Marc Tommasi, Kevin Scaman, Giovanni Neglia
Abstract
This paper presents a new generalization error analysis for Decentralized Stochastic Gradient Descent (D-SGD) based on algorithmic stability. The obtained results overhaul a series of recent works that suggested an increased instability due to decentralization and a detrimental impact of poorly-connected communication graphs on generalization. On the contrary, we show, for convex, strongly convex and non-convex functions, that D-SGD can always recover generalization bounds analogous to those of classical SGD, suggesting that the choice of graph does not matter. We then argue that this result is coming from a worst-case analysis, and we provide a refined optimization-dependent generalization bound for general convex functions. This new bound reveals that the choice of graph can in fact improve the worst-case bound in certain regimes, and that surprisingly, a poorly-connected graph can even be beneficial for generalization.
BibTeX
@inproceedings{
bars2024improved,
title={Improved Stability and Generalization Guarantees of the Decentralized {SGD} Algorithm},
author={Batiste Le bars and Aur{\'e}lien Bellet and Marc Tommasi and Kevin Scaman and Giovanni Neglia},
booktitle={Forty-first International Conference on Machine Learning},
year={2024},
url={https://openreview.net/forum?id=JKPhWzp7Oi}
}