Data Summarization for Federated Learning - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Data Summarization for Federated Learning

Résumé

We explore data summarization techniques as a mean to reduce the energy footprint of Federated Learning (FL). We formulate the problem of selecting a small subset of data points that best represent the gradient of each local dataset as a submodular maximization problem and provide sufficient conditions under which the FL training is guaranteed to converge to the same global model as if the whole local datasets have been used on each client. Experimental results on IID and non-IID datasets show that this approach yields a similar accuracy as training on the full local datasets, but with a significant reduction of runtimes. There is however no clear advantage of data summarization over random sampling.
Fichier principal
Vignette du fichier
MLNreport.pdf (1.17 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04295982 , version 1 (20-11-2023)

Identifiants

  • HAL Id : hal-04295982 , version 1

Citer

Julianna Devillers, Olivier Brun, Balakrishna Prabhu. Data Summarization for Federated Learning. Proceedings of the 6th International Conference on Machine Learning for Networking (MLN'2023), Nov 2023, Paris, France. ⟨hal-04295982⟩
93 Consultations
21 Téléchargements

Partager

Gmail Facebook X LinkedIn More