Tackling Distribution Shifts in Federated Learning with Superquantile Aggregation - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Tackling Distribution Shifts in Federated Learning with Superquantile Aggregation

Résumé

Federated learning has emerged as the predominant framework for distributed machine learning over decentralized data, e.g. in mobile phones. The usual approaches suffer from a distribution shift: the model is trained to fit the average population distribution but is deployed on individual clients, whose data distributions can be quite different. We present a distributionally robust approach to federated learning based on a risk measure known as the superquantile and show how to optimize it by interleaving federated averaging steps with quantile computation. We demonstrate experimentally that our approach is competitive with usual ones in terms of average error and outperforms them in terms of tail statistics of the error.
Fichier principal
Vignette du fichier
main.pdf (1.04 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03834473 , version 1 (30-10-2022)

Identifiants

  • HAL Id : hal-03834473 , version 1

Citer

Krishna Pillutla, Yassine Laguel, Jérôme Malick, Zaid Harchaoui. Tackling Distribution Shifts in Federated Learning with Superquantile Aggregation. NeurIPS 2022 Workshop on Distribution Shifts (DistShift), Dec 2022, New Orleans, United States. ⟨hal-03834473⟩

Collections

CNRS TDS-MACS
55 Consultations
45 Téléchargements

Partager

Gmail Facebook X LinkedIn More