https://scholars.lib.ntu.edu.tw/handle/123456789/629192
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.author | Li, Cen-Jhih | en_US |
dc.contributor.author | Huang, Pin-Han | en_US |
dc.contributor.author | Ma, Yi-Ting | en_US |
dc.contributor.author | HUNG HUNG | en_US |
dc.contributor.author | Huang, Su-Yun | en_US |
dc.date.accessioned | 2023-03-10T06:21:02Z | - |
dc.date.available | 2023-03-10T06:21:02Z | - |
dc.date.issued | 2022-05-13 | - |
dc.identifier.issn | 1099-4300 | - |
dc.identifier.uri | https://scholars.lib.ntu.edu.tw/handle/123456789/629192 | - |
dc.description.abstract | Federated learning is a framework for multiple devices or institutions, called local clients, to collaboratively train a global model without sharing their data. For federated learning with a central server, an aggregation algorithm integrates model information sent from local clients to update the parameters for a global model. Sample mean is the simplest and most commonly used aggregation method. However, it is not robust for data with outliers or under the Byzantine problem, where Byzantine clients send malicious messages to interfere with the learning process. Some robust aggregation methods were introduced in literature including marginal median, geometric median and trimmed-mean. In this article, we propose an alternative robust aggregation method, named γ-mean, which is the minimum divergence estimation based on a robust density power divergence. This γ-mean aggregation mitigates the influence of Byzantine clients by assigning fewer weights. This weighting scheme is data-driven and controlled by the γ value. Robustness from the viewpoint of the influence function is discussed and some numerical results are presented. | en_US |
dc.language.iso | en | en_US |
dc.publisher | MDPI | en_US |
dc.relation.ispartof | Entropy | en_US |
dc.subject | byzantine problem; density power divergence; federated learning; influence function; robustness; γ-divergence | en_US |
dc.title | Robust Aggregation for Federated Learning by Minimum γ-Divergence Estimation | en_US |
dc.type | journal article | en |
dc.identifier.doi | 10.3390/e24050686 | - |
dc.identifier.pmid | 35626569 | - |
dc.identifier.scopus | 2-s2.0-85131306837 | - |
dc.identifier.isi | WOS:000803449900001 | - |
dc.identifier.url | https://api.elsevier.com/content/abstract/scopus_id/85131306837 | - |
dc.relation.journalvolume | 24 | en_US |
dc.relation.journalissue | 5 | en_US |
item.openairetype | journal article | - |
item.openairecristype | http://purl.org/coar/resource_type/c_6501 | - |
item.cerifentitytype | Publications | - |
item.languageiso639-1 | en | - |
item.fulltext | no fulltext | - |
item.grantfulltext | none | - |
crisitem.author.dept | Institute of Health Data Analytics and Statistics | - |
crisitem.author.dept | Public Health | - |
crisitem.author.orcid | 0000-0003-4022-2081 | - |
crisitem.author.parentorg | College of Public Health | - |
crisitem.author.parentorg | College of Public Health | - |
顯示於: | 流行病學與預防醫學研究所 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。