File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Data-Dependent Generalization Bounds for Multi-Class Classification

TitleData-Dependent Generalization Bounds for Multi-Class Classification
Authors
Keywordscovering numbers
Gaussian complexities
generalization error bounds
Multi-class classification
Rademacher complexities
Issue Date2019
Citation
IEEE Transactions on Information Theory, 2019, v. 65, n. 5, p. 2995-3021 How to Cite?
AbstractIn 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 Identifierhttp://hdl.handle.net/10722/329563
ISSN
2023 Impact Factor: 2.2
2023 SCImago Journal Rankings: 1.607
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorLei, Yunwen-
dc.contributor.authorDogan, Urun-
dc.contributor.authorZhou, Ding Xuan-
dc.contributor.authorKloft, Marius-
dc.date.accessioned2023-08-09T03:33:42Z-
dc.date.available2023-08-09T03:33:42Z-
dc.date.issued2019-
dc.identifier.citationIEEE Transactions on Information Theory, 2019, v. 65, n. 5, p. 2995-3021-
dc.identifier.issn0018-9448-
dc.identifier.urihttp://hdl.handle.net/10722/329563-
dc.description.abstractIn 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.languageeng-
dc.relation.ispartofIEEE Transactions on Information Theory-
dc.subjectcovering numbers-
dc.subjectGaussian complexities-
dc.subjectgeneralization error bounds-
dc.subjectMulti-class classification-
dc.subjectRademacher complexities-
dc.titleData-Dependent Generalization Bounds for Multi-Class Classification-
dc.typeArticle-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1109/TIT.2019.2893916-
dc.identifier.scopuseid_2-s2.0-85065143828-
dc.identifier.volume65-
dc.identifier.issue5-
dc.identifier.spage2995-
dc.identifier.epage3021-
dc.identifier.eissn1557-9654-
dc.identifier.isiWOS:000466029900023-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats