File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1016/j.jfranklin.2018.11.021
- Scopus: eid_2-s2.0-85060958164
- WOS: WOS:000460043800021
- Find via
Supplementary
- Citations:
- Appears in Collections:
Article: Calibration of ϵ−insensitive loss in support vector machines regression
Title | Calibration of ϵ−insensitive loss in support vector machines regression |
---|---|
Authors | |
Issue Date | 2019 |
Citation | Journal of the Franklin Institute, 2019, v. 356, n. 4, p. 2111-2129 How to Cite? |
Abstract | © 2018 The Franklin Institute Support vector machines regression (SVMR) is an important tool in many machine learning applications. In this paper, we focus on the theoretical understanding of SVMR based on the ϵ−insensitive loss. For fixed ϵ ≥ 0 and general data generating distributions, we show that the minimizer of the expected risk for ϵ−insensitive loss used in SVMR is a set-valued function called conditional ϵ−median. We then establish a calibration inequality of ϵ−insensitive loss under a noise condition on the conditional distributions. This inequality also ensures us to present a nontrivial variance-expectation bound for ϵ−insensitive loss, and which is known to be important in statistical analysis of the regularized learning algorithms. With the help of the calibration inequality and variance-expectation bound, we finally derive an explicit learning rate for SVMR in some L r −space. |
Persistent Identifier | http://hdl.handle.net/10722/276629 |
ISSN | 2023 Impact Factor: 3.7 2023 SCImago Journal Rankings: 1.191 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Tong, Hongzhi | - |
dc.contributor.author | Ng, Michael K. | - |
dc.date.accessioned | 2019-09-18T08:34:11Z | - |
dc.date.available | 2019-09-18T08:34:11Z | - |
dc.date.issued | 2019 | - |
dc.identifier.citation | Journal of the Franklin Institute, 2019, v. 356, n. 4, p. 2111-2129 | - |
dc.identifier.issn | 0016-0032 | - |
dc.identifier.uri | http://hdl.handle.net/10722/276629 | - |
dc.description.abstract | © 2018 The Franklin Institute Support vector machines regression (SVMR) is an important tool in many machine learning applications. In this paper, we focus on the theoretical understanding of SVMR based on the ϵ−insensitive loss. For fixed ϵ ≥ 0 and general data generating distributions, we show that the minimizer of the expected risk for ϵ−insensitive loss used in SVMR is a set-valued function called conditional ϵ−median. We then establish a calibration inequality of ϵ−insensitive loss under a noise condition on the conditional distributions. This inequality also ensures us to present a nontrivial variance-expectation bound for ϵ−insensitive loss, and which is known to be important in statistical analysis of the regularized learning algorithms. With the help of the calibration inequality and variance-expectation bound, we finally derive an explicit learning rate for SVMR in some L r −space. | - |
dc.language | eng | - |
dc.relation.ispartof | Journal of the Franklin Institute | - |
dc.title | Calibration of ϵ−insensitive loss in support vector machines regression | - |
dc.type | Article | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1016/j.jfranklin.2018.11.021 | - |
dc.identifier.scopus | eid_2-s2.0-85060958164 | - |
dc.identifier.volume | 356 | - |
dc.identifier.issue | 4 | - |
dc.identifier.spage | 2111 | - |
dc.identifier.epage | 2129 | - |
dc.identifier.isi | WOS:000460043800021 | - |
dc.identifier.issnl | 0016-0032 | - |