DC Field | Value | Language |
---|---|---|
dc.contributor.author | Lee, Jongmyeong | ko |
dc.contributor.author | Jang, Youngsu | ko |
dc.contributor.author | Lee, Jangho | ko |
dc.contributor.author | Kang, Joonhyuk | ko |
dc.date.accessioned | 2022-12-05T01:01:05Z | - |
dc.date.available | 2022-12-05T01:01:05Z | - |
dc.date.created | 2022-12-02 | - |
dc.date.created | 2022-12-02 | - |
dc.date.issued | 2022-10-20 | - |
dc.identifier.citation | The 13th International Conference on ICT Convergence, ICTC 2022, pp.2044 - 2047 | - |
dc.identifier.uri | http://hdl.handle.net/10203/301578 | - |
dc.description.abstract | High communication overhead is a major bottleneck in federated learning (FL). To overcome this issue, sparsification is utilized in various compression frameworks. Generally, local clients upload the updated weights to the server. However, in sparsification, we observed that local clients upload the difference between the updated weights and the original weights. Our study is to confirm the importance of uploading the difference of weights in sparsification and to observe how different the accuracy between the two schemes is. | - |
dc.language | English | - |
dc.publisher | Institute of Electrical and Electronics Engineers Inc. | - |
dc.title | Sparsification on Different Federated Learning Schemes: Comparative Analysis | - |
dc.type | Conference | - |
dc.identifier.scopusid | 2-s2.0-85143256431 | - |
dc.type.rims | CONF | - |
dc.citation.beginningpage | 2044 | - |
dc.citation.endingpage | 2047 | - |
dc.citation.publicationname | The 13th International Conference on ICT Convergence, ICTC 2022 | - |
dc.identifier.conferencecountry | KO | - |
dc.identifier.conferencelocation | Ramada Plaza Hotel Jeju & Online | - |
dc.identifier.doi | 10.1109/ICTC55196.2022.9952431 | - |
dc.contributor.localauthor | Kang, Joonhyuk | - |
dc.contributor.nonIdAuthor | Lee, Jongmyeong | - |
dc.contributor.nonIdAuthor | Lee, Jangho | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.