An Experimental Study of Different Aggregation Schemes in Semi-Asynchronous Federated Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Yunbo, Gui, Jiaping, Wu, Yue
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913362767511552
author Li, Yunbo
Gui, Jiaping
Wu, Yue
author_facet Li, Yunbo
Gui, Jiaping
Wu, Yue
contents Federated learning is highly valued due to its high-performance computing in distributed environments while safeguarding data privacy. To address resource heterogeneity, researchers have proposed a semi-asynchronous federated learning (SAFL) architecture. However, the performance gap between different aggregation targets in SAFL remain unexplored. In this paper, we systematically compare the performance between two algorithm modes, FedSGD and FedAvg that correspond to aggregating gradients and models, respectively. Our results across various task scenarios indicate these two modes exhibit a substantial performance gap. Specifically, FedSGD achieves higher accuracy and faster convergence but experiences more severe fluctuates in accuracy, whereas FedAvg excels in handling straggler issues but converges slower with reduced accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2405_16086
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle An Experimental Study of Different Aggregation Schemes in Semi-Asynchronous Federated Learning
Li, Yunbo
Gui, Jiaping
Wu, Yue
Distributed, Parallel, and Cluster Computing
Performance
Federated learning is highly valued due to its high-performance computing in distributed environments while safeguarding data privacy. To address resource heterogeneity, researchers have proposed a semi-asynchronous federated learning (SAFL) architecture. However, the performance gap between different aggregation targets in SAFL remain unexplored. In this paper, we systematically compare the performance between two algorithm modes, FedSGD and FedAvg that correspond to aggregating gradients and models, respectively. Our results across various task scenarios indicate these two modes exhibit a substantial performance gap. Specifically, FedSGD achieves higher accuracy and faster convergence but experiences more severe fluctuates in accuracy, whereas FedAvg excels in handling straggler issues but converges slower with reduced accuracy.
title An Experimental Study of Different Aggregation Schemes in Semi-Asynchronous Federated Learning
topic Distributed, Parallel, and Cluster Computing
Performance
url https://arxiv.org/abs/2405.16086