File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1109/TIT.2019.2893916
- Scopus: eid_2-s2.0-85065143828
- WOS: WOS:000466029900023
- Find via
Supplementary
- Citations:
- Appears in Collections:
Article: Data-Dependent Generalization Bounds for Multi-Class Classification
Title | Data-Dependent Generalization Bounds for Multi-Class Classification |
---|---|
Authors | |
Keywords | covering numbers Gaussian complexities generalization error bounds Multi-class classification Rademacher complexities |
Issue Date | 2019 |
Citation | IEEE Transactions on Information Theory, 2019, v. 65, n. 5, p. 2995-3021 How to Cite? |
Abstract | In this paper, we study data-dependent generalization error bounds that exhibit a mild dependency on the number of classes, making them suitable for multi-class learning with a large number of label classes. The bounds generally hold for empirical multi-class risk minimization algorithms using an arbitrary norm as the regularizer. Key to our analysis is new structural results for multi-class Gaussian complexities and empirical $\ell -\infty $ -norm covering numbers, which exploit the Lipschitz continuity of the loss function with respect to the $\ell -{2}$ - and $\ell -\infty $ -norm, respectively. We establish data-dependent error bounds in terms of the complexities of a linear function class defined on a finite set induced by training examples, for which we show tight lower and upper bounds. We apply the results to several prominent multi-class learning machines and show a tighter dependency on the number of classes than the state of the art. For instance, for the multi-class support vector machine of Crammer and Singer (2002), we obtain a data-dependent bound with a logarithmic dependency, which is a significant improvement of the previous square-root dependency. The experimental results are reported to verify the effectiveness of our theoretical findings. |
Persistent Identifier | http://hdl.handle.net/10722/329563 |
ISSN | 2023 Impact Factor: 2.2 2023 SCImago Journal Rankings: 1.607 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Lei, Yunwen | - |
dc.contributor.author | Dogan, Urun | - |
dc.contributor.author | Zhou, Ding Xuan | - |
dc.contributor.author | Kloft, Marius | - |
dc.date.accessioned | 2023-08-09T03:33:42Z | - |
dc.date.available | 2023-08-09T03:33:42Z | - |
dc.date.issued | 2019 | - |
dc.identifier.citation | IEEE Transactions on Information Theory, 2019, v. 65, n. 5, p. 2995-3021 | - |
dc.identifier.issn | 0018-9448 | - |
dc.identifier.uri | http://hdl.handle.net/10722/329563 | - |
dc.description.abstract | In this paper, we study data-dependent generalization error bounds that exhibit a mild dependency on the number of classes, making them suitable for multi-class learning with a large number of label classes. The bounds generally hold for empirical multi-class risk minimization algorithms using an arbitrary norm as the regularizer. Key to our analysis is new structural results for multi-class Gaussian complexities and empirical $\ell -\infty $ -norm covering numbers, which exploit the Lipschitz continuity of the loss function with respect to the $\ell -{2}$ - and $\ell -\infty $ -norm, respectively. We establish data-dependent error bounds in terms of the complexities of a linear function class defined on a finite set induced by training examples, for which we show tight lower and upper bounds. We apply the results to several prominent multi-class learning machines and show a tighter dependency on the number of classes than the state of the art. For instance, for the multi-class support vector machine of Crammer and Singer (2002), we obtain a data-dependent bound with a logarithmic dependency, which is a significant improvement of the previous square-root dependency. The experimental results are reported to verify the effectiveness of our theoretical findings. | - |
dc.language | eng | - |
dc.relation.ispartof | IEEE Transactions on Information Theory | - |
dc.subject | covering numbers | - |
dc.subject | Gaussian complexities | - |
dc.subject | generalization error bounds | - |
dc.subject | Multi-class classification | - |
dc.subject | Rademacher complexities | - |
dc.title | Data-Dependent Generalization Bounds for Multi-Class Classification | - |
dc.type | Article | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1109/TIT.2019.2893916 | - |
dc.identifier.scopus | eid_2-s2.0-85065143828 | - |
dc.identifier.volume | 65 | - |
dc.identifier.issue | 5 | - |
dc.identifier.spage | 2995 | - |
dc.identifier.epage | 3021 | - |
dc.identifier.eissn | 1557-9654 | - |
dc.identifier.isi | WOS:000466029900023 | - |