Node selection based on label quantity information in federated learning

Aiming at the problem that the difference of node data distribution has adverse effect on the performance of federated learning algorithm, a node selection algorithm based on label quantity information was proposed.An optimization objective based on the label quantity information of nodes was design...

Full description

Saved in:
Bibliographic Details
Main Authors: Jiahua MA, Xinghua SUN, Wenchao XIA, Xijun WANG, Hongzhou TAN, Hongbo ZHU
Format: Article
Language:zho
Published: China InfoCom Media Group 2021-12-01
Series:物联网学报
Subjects:
Online Access:http://www.wlwxb.com.cn/zh/article/doi/10.11959/j.issn.2096-3750.2021.00249/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850091095443111936
author Jiahua MA
Xinghua SUN
Wenchao XIA
Xijun WANG
Hongzhou TAN
Hongbo ZHU
author_facet Jiahua MA
Xinghua SUN
Wenchao XIA
Xijun WANG
Hongzhou TAN
Hongbo ZHU
author_sort Jiahua MA
collection DOAJ
description Aiming at the problem that the difference of node data distribution has adverse effect on the performance of federated learning algorithm, a node selection algorithm based on label quantity information was proposed.An optimization objective based on the label quantity information of nodes was designed, considering the optimization problem of selecting the nodes with balanced label distribution under a certain time consumption limit.According to the correlation between the aggregated label distribution of selected nodes and the convergence of the global model, the upper bound of the weight divergence of the global model was reduced to improve the convergence stability of the algorithm.Simulation results shows that the new algorithm had higher convergence efficiency than the existing node selection algorithm.
format Article
id doaj-art-748e3400ad4f43ab92d918b211dcc498
institution DOAJ
issn 2096-3750
language zho
publishDate 2021-12-01
publisher China InfoCom Media Group
record_format Article
series 物联网学报
spelling doaj-art-748e3400ad4f43ab92d918b211dcc4982025-08-20T02:42:26ZzhoChina InfoCom Media Group物联网学报2096-37502021-12-015465359647552Node selection based on label quantity information in federated learningJiahua MAXinghua SUNWenchao XIAXijun WANGHongzhou TANHongbo ZHUAiming at the problem that the difference of node data distribution has adverse effect on the performance of federated learning algorithm, a node selection algorithm based on label quantity information was proposed.An optimization objective based on the label quantity information of nodes was designed, considering the optimization problem of selecting the nodes with balanced label distribution under a certain time consumption limit.According to the correlation between the aggregated label distribution of selected nodes and the convergence of the global model, the upper bound of the weight divergence of the global model was reduced to improve the convergence stability of the algorithm.Simulation results shows that the new algorithm had higher convergence efficiency than the existing node selection algorithm.http://www.wlwxb.com.cn/zh/article/doi/10.11959/j.issn.2096-3750.2021.00249/federated learningnode selectioncommunication delay
spellingShingle Jiahua MA
Xinghua SUN
Wenchao XIA
Xijun WANG
Hongzhou TAN
Hongbo ZHU
Node selection based on label quantity information in federated learning
物联网学报
federated learning
node selection
communication delay
title Node selection based on label quantity information in federated learning
title_full Node selection based on label quantity information in federated learning
title_fullStr Node selection based on label quantity information in federated learning
title_full_unstemmed Node selection based on label quantity information in federated learning
title_short Node selection based on label quantity information in federated learning
title_sort node selection based on label quantity information in federated learning
topic federated learning
node selection
communication delay
url http://www.wlwxb.com.cn/zh/article/doi/10.11959/j.issn.2096-3750.2021.00249/
work_keys_str_mv AT jiahuama nodeselectionbasedonlabelquantityinformationinfederatedlearning
AT xinghuasun nodeselectionbasedonlabelquantityinformationinfederatedlearning
AT wenchaoxia nodeselectionbasedonlabelquantityinformationinfederatedlearning
AT xijunwang nodeselectionbasedonlabelquantityinformationinfederatedlearning
AT hongzhoutan nodeselectionbasedonlabelquantityinformationinfederatedlearning
AT hongbozhu nodeselectionbasedonlabelquantityinformationinfederatedlearning