A DC programming approach for feature selection in the Minimax Probability Machine
This paper presents a new feature selection framework based on the -norm, in which data are summarized by their moments of the class conditional densities. However, discontinuity of the -norm makes it difficult to find the optimal solution. We apply a proper approximation of the -norm and a bound on...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2014-01-01
|
| Series: | International Journal of Computational Intelligence Systems |
| Subjects: | |
| Online Access: | https://www.atlantis-press.com/article/25868468.pdf |
| _version_ | 1818233660688564224 |
|---|---|
| author | Liming Yang Ribo Ju |
| author_facet | Liming Yang Ribo Ju |
| author_sort | Liming Yang |
| collection | DOAJ |
| description | This paper presents a new feature selection framework based on the -norm, in which data are summarized by their moments of the class conditional densities. However, discontinuity of the -norm makes it difficult to find the optimal solution. We apply a proper approximation of the -norm and a bound on the misclassification probability involving the mean and covariance of the dataset, to derive a robust difference of convex functions (DC) program formulation, while the DC optimization algorithm is used to solve the problem effectively. Furthermore, a kernelized version of this problem is also presented in this work. Experimental results on both real and synthetic datasets show that the proposed formulations can select fewer features than the traditional Minimax Probability Machine and the -norm state. |
| first_indexed | 2024-12-12T11:25:43Z |
| format | Article |
| id | doaj.art-cf3f577ec4404e3f91aa7eae2c89ea52 |
| institution | Directory Open Access Journal |
| issn | 1875-6883 |
| language | English |
| last_indexed | 2024-12-12T11:25:43Z |
| publishDate | 2014-01-01 |
| publisher | Springer |
| record_format | Article |
| series | International Journal of Computational Intelligence Systems |
| spelling | doaj.art-cf3f577ec4404e3f91aa7eae2c89ea522022-12-22T00:25:56ZengSpringerInternational Journal of Computational Intelligence Systems1875-68832014-01-017110.1080/18756891.2013.864471A DC programming approach for feature selection in the Minimax Probability MachineLiming YangRibo JuThis paper presents a new feature selection framework based on the -norm, in which data are summarized by their moments of the class conditional densities. However, discontinuity of the -norm makes it difficult to find the optimal solution. We apply a proper approximation of the -norm and a bound on the misclassification probability involving the mean and covariance of the dataset, to derive a robust difference of convex functions (DC) program formulation, while the DC optimization algorithm is used to solve the problem effectively. Furthermore, a kernelized version of this problem is also presented in this work. Experimental results on both real and synthetic datasets show that the proposed formulations can select fewer features than the traditional Minimax Probability Machine and the -norm state.https://www.atlantis-press.com/article/25868468.pdfFeature selectionMinimax probability machineDC programming |
| spellingShingle | Liming Yang Ribo Ju A DC programming approach for feature selection in the Minimax Probability Machine International Journal of Computational Intelligence Systems Feature selection Minimax probability machine DC programming |
| title | A DC programming approach for feature selection in the Minimax Probability Machine |
| title_full | A DC programming approach for feature selection in the Minimax Probability Machine |
| title_fullStr | A DC programming approach for feature selection in the Minimax Probability Machine |
| title_full_unstemmed | A DC programming approach for feature selection in the Minimax Probability Machine |
| title_short | A DC programming approach for feature selection in the Minimax Probability Machine |
| title_sort | dc programming approach for feature selection in the minimax probability machine |
| topic | Feature selection Minimax probability machine DC programming |
| url | https://www.atlantis-press.com/article/25868468.pdf |
| work_keys_str_mv | AT limingyang adcprogrammingapproachforfeatureselectionintheminimaxprobabilitymachine AT riboju adcprogrammingapproachforfeatureselectionintheminimaxprobabilitymachine AT limingyang dcprogrammingapproachforfeatureselectionintheminimaxprobabilitymachine AT riboju dcprogrammingapproachforfeatureselectionintheminimaxprobabilitymachine |