Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Mora, Alessio, Tenison, Irene, Bellavista, Paolo, Rish, Irina
Format: Preprint
Veröffentlicht: 2022
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2211.04742
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916675311370240
author Mora, Alessio
Tenison, Irene
Bellavista, Paolo
Rish, Irina
author_facet Mora, Alessio
Tenison, Irene
Bellavista, Paolo
Rish, Irina
contents Federated Learning (FL) enables the training of Deep Learning models without centrally collecting possibly sensitive raw data. The most used algorithms for FL are parameter-averaging based schemes (e.g., Federated Averaging) that, however, have well known limits, i.e., model homogeneity, high communication cost, poor performance in presence of heterogeneous data distributions. Federated adaptations of regular Knowledge Distillation (KD) can solve or mitigate the weaknesses of parameter-averaging FL algorithms while possibly introducing other trade-offs. In this article, we originally present a focused review of the state-of-the-art KD-based algorithms specifically tailored for FL, by providing both a novel classification of the existing approaches and a detailed technical description of their pros, cons, and tradeoffs.
format Preprint
id arxiv_https___arxiv_org_abs_2211_04742
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Knowledge Distillation for Federated Learning: a Practical Guide
Mora, Alessio
Tenison, Irene
Bellavista, Paolo
Rish, Irina
Machine Learning
Federated Learning (FL) enables the training of Deep Learning models without centrally collecting possibly sensitive raw data. The most used algorithms for FL are parameter-averaging based schemes (e.g., Federated Averaging) that, however, have well known limits, i.e., model homogeneity, high communication cost, poor performance in presence of heterogeneous data distributions. Federated adaptations of regular Knowledge Distillation (KD) can solve or mitigate the weaknesses of parameter-averaging FL algorithms while possibly introducing other trade-offs. In this article, we originally present a focused review of the state-of-the-art KD-based algorithms specifically tailored for FL, by providing both a novel classification of the existing approaches and a detailed technical description of their pros, cons, and tradeoffs.
title Knowledge Distillation for Federated Learning: a Practical Guide
topic Machine Learning
url https://arxiv.org/abs/2211.04742