Sporadic Gradient Tracking over Directed Graphs: A Theoretical Perspective on Decentralized Federated Learning
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914298528268288 |
|---|---|
| author | Zehtabi, Shahryar Han, Dong-Jun Hosseinalipour, Seyyedali Brinton, Christopher |
| author_facet | Zehtabi, Shahryar Han, Dong-Jun Hosseinalipour, Seyyedali Brinton, Christopher |
| contents | Decentralized Federated Learning (DFL) enables clients with local data to collaborate in a peer-to-peer manner to train a generalized model. In this paper, we unify two branches of work that have separately solved important challenges in DFL: (i) gradient tracking techniques for mitigating data heterogeneity and (ii) accounting for diverse availability of resources across clients. We propose $\textit{Sporadic Gradient Tracking}$ ($\texttt{Spod-GT}$), the first DFL algorithm that incorporates these factors over general directed graphs by allowing (i) client-specific gradient computation frequencies and (ii) heterogeneous and asymmetric communication frequencies. We conduct a rigorous convergence analysis of our methodology with relaxed assumptions on gradient estimation variance and gradient diversity of clients, providing consensus and optimality guarantees for GT over directed graphs despite intermittent client participation. Through numerical experiments on image classification datasets, we demonstrate the efficacy of $\texttt{Spod-GT}$ compared to well-known GT baselines. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_00791 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Sporadic Gradient Tracking over Directed Graphs: A Theoretical Perspective on Decentralized Federated Learning Zehtabi, Shahryar Han, Dong-Jun Hosseinalipour, Seyyedali Brinton, Christopher Machine Learning Optimization and Control Decentralized Federated Learning (DFL) enables clients with local data to collaborate in a peer-to-peer manner to train a generalized model. In this paper, we unify two branches of work that have separately solved important challenges in DFL: (i) gradient tracking techniques for mitigating data heterogeneity and (ii) accounting for diverse availability of resources across clients. We propose $\textit{Sporadic Gradient Tracking}$ ($\texttt{Spod-GT}$), the first DFL algorithm that incorporates these factors over general directed graphs by allowing (i) client-specific gradient computation frequencies and (ii) heterogeneous and asymmetric communication frequencies. We conduct a rigorous convergence analysis of our methodology with relaxed assumptions on gradient estimation variance and gradient diversity of clients, providing consensus and optimality guarantees for GT over directed graphs despite intermittent client participation. Through numerical experiments on image classification datasets, we demonstrate the efficacy of $\texttt{Spod-GT}$ compared to well-known GT baselines. |
| title | Sporadic Gradient Tracking over Directed Graphs: A Theoretical Perspective on Decentralized Federated Learning |
| topic | Machine Learning Optimization and Control |
| url | https://arxiv.org/abs/2602.00791 |