A feature selection method for multivariate performance measures
Feature selection with specific multivariate performance measures is the key to the success of many applications such as image retrieval and text classification. The existing feature selection methods are usually designed for classification error. In this paper, we propose a generalized sparse regul...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/100979 http://hdl.handle.net/10220/16693 |
| _version_ | 1811694354939510784 |
|---|---|
| author | Mao, Qi Tsang, Ivor Wai-Hung |
| author2 | School of Computer Engineering |
| author_facet | School of Computer Engineering Mao, Qi Tsang, Ivor Wai-Hung |
| author_sort | Mao, Qi |
| collection | NTU |
| description | Feature selection with specific multivariate performance measures is the key to the success of many applications such as image retrieval and text classification. The existing feature selection methods are usually designed for classification error. In this paper, we propose a generalized sparse regularizer. Based on the proposed regularizer, we present a unified feature selection framework for general loss functions. In particular, we study the novel feature selection paradigm by optimizing multivariate performance measures. The resultant formulation is a challenging problem for high-dimensional data. Hence, a two-layer cutting plane algorithm is proposed to solve this problem, and the convergence is presented. In addition, we adapt the proposed method to optimize multivariate measures for multiple-instance learning problems. The analyses by comparing with the state-of-the-art feature selection methods show that the proposed method is superior to others. Extensive experiments on large-scale and high-dimensional real-world datasets show that the proposed method outperforms l1-SVM and SVM-RFE when choosing a small subset of features, and achieves significantly improved performances over SVMperl in terms of F1-score. |
| first_indexed | 2024-10-01T07:06:15Z |
| format | Journal Article |
| id | ntu-10356/100979 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T07:06:15Z |
| publishDate | 2013 |
| record_format | dspace |
| spelling | ntu-10356/1009792020-05-28T07:17:43Z A feature selection method for multivariate performance measures Mao, Qi Tsang, Ivor Wai-Hung School of Computer Engineering DRNTU::Engineering::Computer science and engineering Feature selection with specific multivariate performance measures is the key to the success of many applications such as image retrieval and text classification. The existing feature selection methods are usually designed for classification error. In this paper, we propose a generalized sparse regularizer. Based on the proposed regularizer, we present a unified feature selection framework for general loss functions. In particular, we study the novel feature selection paradigm by optimizing multivariate performance measures. The resultant formulation is a challenging problem for high-dimensional data. Hence, a two-layer cutting plane algorithm is proposed to solve this problem, and the convergence is presented. In addition, we adapt the proposed method to optimize multivariate measures for multiple-instance learning problems. The analyses by comparing with the state-of-the-art feature selection methods show that the proposed method is superior to others. Extensive experiments on large-scale and high-dimensional real-world datasets show that the proposed method outperforms l1-SVM and SVM-RFE when choosing a small subset of features, and achieves significantly improved performances over SVMperl in terms of F1-score. 2013-10-23T04:43:01Z 2019-12-06T20:31:42Z 2013-10-23T04:43:01Z 2019-12-06T20:31:42Z 2013 2013 Journal Article Mao, Q., & Tsang, I. W. H. (2013). A feature selection method for multivariate performance measures. IEEE Transactions on Pattern Analysis and Machine Intelligence, 35(9), 2051-2063. 0162-8828 https://hdl.handle.net/10356/100979 http://hdl.handle.net/10220/16693 10.1109/TPAMI.2012.266 en IEEE Transactions on Pattern Analysis and Machine Intelligence |
| spellingShingle | DRNTU::Engineering::Computer science and engineering Mao, Qi Tsang, Ivor Wai-Hung A feature selection method for multivariate performance measures |
| title | A feature selection method for multivariate performance measures |
| title_full | A feature selection method for multivariate performance measures |
| title_fullStr | A feature selection method for multivariate performance measures |
| title_full_unstemmed | A feature selection method for multivariate performance measures |
| title_short | A feature selection method for multivariate performance measures |
| title_sort | feature selection method for multivariate performance measures |
| topic | DRNTU::Engineering::Computer science and engineering |
| url | https://hdl.handle.net/10356/100979 http://hdl.handle.net/10220/16693 |
| work_keys_str_mv | AT maoqi afeatureselectionmethodformultivariateperformancemeasures AT tsangivorwaihung afeatureselectionmethodformultivariateperformancemeasures AT maoqi featureselectionmethodformultivariateperformancemeasures AT tsangivorwaihung featureselectionmethodformultivariateperformancemeasures |