Saved in:
Bibliographic Details
Main Authors: Jiang, Enyi, Zhang, Yibo Jacky, Koyejo, Sanmi
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2302.05049
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929286781337600
author Jiang, Enyi
Zhang, Yibo Jacky
Koyejo, Sanmi
author_facet Jiang, Enyi
Zhang, Yibo Jacky
Koyejo, Sanmi
contents Federated Domain Adaptation (FDA) describes the federated learning (FL) setting where source clients and a server work collaboratively to improve the performance of a target client where limited data is available. The domain shift between the source and target domains, coupled with limited data of the target client, makes FDA a challenging problem, e.g., common techniques such as federated averaging and fine-tuning fail due to domain shift and data scarcity. To theoretically understand the problem, we introduce new metrics that characterize the FDA setting and a theoretical framework with novel theorems for analyzing the performance of server aggregation rules. Further, we propose a novel lightweight aggregation rule, Federated Gradient Projection ($\texttt{FedGP}$), which significantly improves the target performance with domain shift and data scarcity. Moreover, our theory suggests an $\textit{auto-weighting scheme}$ that finds the optimal combinations of the source and target gradients. This scheme improves both $\texttt{FedGP}$ and a simpler heuristic aggregation rule. Extensive experiments verify the theoretical insights and illustrate the effectiveness of the proposed methods in practice.
format Preprint
id arxiv_https___arxiv_org_abs_2302_05049
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Principled Federated Domain Adaptation: Gradient Projection and Auto-Weighting
Jiang, Enyi
Zhang, Yibo Jacky
Koyejo, Sanmi
Machine Learning
Federated Domain Adaptation (FDA) describes the federated learning (FL) setting where source clients and a server work collaboratively to improve the performance of a target client where limited data is available. The domain shift between the source and target domains, coupled with limited data of the target client, makes FDA a challenging problem, e.g., common techniques such as federated averaging and fine-tuning fail due to domain shift and data scarcity. To theoretically understand the problem, we introduce new metrics that characterize the FDA setting and a theoretical framework with novel theorems for analyzing the performance of server aggregation rules. Further, we propose a novel lightweight aggregation rule, Federated Gradient Projection ($\texttt{FedGP}$), which significantly improves the target performance with domain shift and data scarcity. Moreover, our theory suggests an $\textit{auto-weighting scheme}$ that finds the optimal combinations of the source and target gradients. This scheme improves both $\texttt{FedGP}$ and a simpler heuristic aggregation rule. Extensive experiments verify the theoretical insights and illustrate the effectiveness of the proposed methods in practice.
title Principled Federated Domain Adaptation: Gradient Projection and Auto-Weighting
topic Machine Learning
url https://arxiv.org/abs/2302.05049