Contractive rectifier networks for nonlinear maximum margin classification
© 2015 IEEE. To find the optimal nonlinear separating boundary with maximum margin in the input data space, this paper proposes Contractive Rectifier Networks (CRNs), wherein the hidden-layer transformations are restricted to be contraction mappings. The contractive constraints ensure that the achie...
| Main Authors: | , , , , , |
|---|---|
| Format: | Conference Paper |
| Published: |
2015
|
| Online Access: | http://hdl.handle.net/20.500.11937/69913 |
| _version_ | 1848762165232664576 |
|---|---|
| author | An, Senjian Hayat, M. Khan, S. Bennamoun, M. Boussaid, F. Sohel, F. |
| author_facet | An, Senjian Hayat, M. Khan, S. Bennamoun, M. Boussaid, F. Sohel, F. |
| author_sort | An, Senjian |
| building | Curtin Institutional Repository |
| collection | Online Access |
| description | © 2015 IEEE. To find the optimal nonlinear separating boundary with maximum margin in the input data space, this paper proposes Contractive Rectifier Networks (CRNs), wherein the hidden-layer transformations are restricted to be contraction mappings. The contractive constraints ensure that the achieved separating margin in the input space is larger than or equal to the separating margin in the output layer. The training of the proposed CRNs is formulated as a linear support vector machine (SVM) in the output layer, combined with two or more contractive hidden layers. Effective algorithms have been proposed to address the optimization challenges arising from contraction constraints. Experimental results on MNIST, CIFAR-10, CIFAR-100 and MIT-67 datasets demonstrate that the proposed contractive rectifier networks consistently outperform their conventional unconstrained rectifier network counterparts. |
| first_indexed | 2025-11-14T10:43:14Z |
| format | Conference Paper |
| id | curtin-20.500.11937-69913 |
| institution | Curtin University Malaysia |
| institution_category | Local University |
| last_indexed | 2025-11-14T10:43:14Z |
| publishDate | 2015 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | curtin-20.500.11937-699132018-08-08T04:57:21Z Contractive rectifier networks for nonlinear maximum margin classification An, Senjian Hayat, M. Khan, S. Bennamoun, M. Boussaid, F. Sohel, F. © 2015 IEEE. To find the optimal nonlinear separating boundary with maximum margin in the input data space, this paper proposes Contractive Rectifier Networks (CRNs), wherein the hidden-layer transformations are restricted to be contraction mappings. The contractive constraints ensure that the achieved separating margin in the input space is larger than or equal to the separating margin in the output layer. The training of the proposed CRNs is formulated as a linear support vector machine (SVM) in the output layer, combined with two or more contractive hidden layers. Effective algorithms have been proposed to address the optimization challenges arising from contraction constraints. Experimental results on MNIST, CIFAR-10, CIFAR-100 and MIT-67 datasets demonstrate that the proposed contractive rectifier networks consistently outperform their conventional unconstrained rectifier network counterparts. 2015 Conference Paper http://hdl.handle.net/20.500.11937/69913 10.1109/ICCV.2015.289 restricted |
| spellingShingle | An, Senjian Hayat, M. Khan, S. Bennamoun, M. Boussaid, F. Sohel, F. Contractive rectifier networks for nonlinear maximum margin classification |
| title | Contractive rectifier networks for nonlinear maximum margin classification |
| title_full | Contractive rectifier networks for nonlinear maximum margin classification |
| title_fullStr | Contractive rectifier networks for nonlinear maximum margin classification |
| title_full_unstemmed | Contractive rectifier networks for nonlinear maximum margin classification |
| title_short | Contractive rectifier networks for nonlinear maximum margin classification |
| title_sort | contractive rectifier networks for nonlinear maximum margin classification |
| url | http://hdl.handle.net/20.500.11937/69913 |