DC Field | Value | Language |
---|---|---|
dc.contributor.author | Sohn, Jy-yong | ko |
dc.contributor.author | Shang, Liang | ko |
dc.contributor.author | Chen, Hongxu | ko |
dc.contributor.author | Moon, Jaekyun | ko |
dc.contributor.author | Papailiopoulos, Dimitris | ko |
dc.contributor.author | Lee, Kangwook | ko |
dc.date.accessioned | 2022-11-28T08:04:01Z | - |
dc.date.available | 2022-11-28T08:04:01Z | - |
dc.date.created | 2022-11-28 | - |
dc.date.created | 2022-11-28 | - |
dc.date.created | 2022-11-28 | - |
dc.date.issued | 2022-07-23 | - |
dc.identifier.citation | The 39th International Conference on Machine Learning, ICML 2022 | - |
dc.identifier.issn | 2640-3498 | - |
dc.identifier.uri | http://hdl.handle.net/10203/301174 | - |
dc.description.abstract | Mixup is a data augmentation method that generates new data points by mixing a pair of input data. While mixup generally improves the prediction performance, it sometimes degrades the performance. In this paper, we first identify the main causes of this phenomenon by theoretically and empirically analyzing the mixup algorithm. To resolve this, we propose GenLabel, a simple yet effective relabeling algorithm designed for mixup. In particular, GenLabel helps the mixup algorithm correctly label mixup samples by learning the class-conditional data distribution using generative models. Via theoretical and empirical analysis, we show that mixup, when used together with GenLabel, can effectively resolve the aforementioned phenomenon, improving the accuracy of mixup-trained model. | - |
dc.language | English | - |
dc.publisher | International Conference on Machine Learning | - |
dc.title | GenLabel: Mixup Relabeling using Generative Models | - |
dc.type | Conference | - |
dc.identifier.wosid | 000900130201020 | - |
dc.identifier.scopusid | 2-s2.0-85163066692 | - |
dc.type.rims | CONF | - |
dc.citation.publicationname | The 39th International Conference on Machine Learning, ICML 2022 | - |
dc.identifier.conferencecountry | US | - |
dc.identifier.conferencelocation | Baltimore, Maryland | - |
dc.contributor.localauthor | Moon, Jaekyun | - |
dc.contributor.nonIdAuthor | Sohn, Jy-yong | - |
dc.contributor.nonIdAuthor | Shang, Liang | - |
dc.contributor.nonIdAuthor | Chen, Hongxu | - |
dc.contributor.nonIdAuthor | Papailiopoulos, Dimitris | - |
dc.contributor.nonIdAuthor | Lee, Kangwook | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.