Distance metric learning : a two-phase approach
- Author
- Bac Nguyen Cong (UGent) , Carlos Morell and Bernard De Baets (UGent)
- Organization
- Abstract
- Distance metric learning has been successfully incorporated in many machine learning applications. The main challenge arises from the positive semidefiniteness constraint on the Mahalanobis matrix, which results in a high computational cost. In this paper, we develop a novel approach to reduce this computational burden. We first map each training example into a new space by an orthonormal transformation. Then, in the transformed space, we simply learn a diagonal matrix. This two-phase approach is thus much easier and less costly than learning a full Mahalanobis matrix in one phase as is commonly done.
Downloads
-
es2017-8.pdf
- full text (Published version)
- |
- open access
- |
- |
- 1.50 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-8532218
- MLA
- Nguyen Cong, Bac, et al. “Distance Metric Learning : A Two-Phase Approach.” Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2017), ESANN, 2017, pp. 123–28.
- APA
- Nguyen Cong, B., Morell, C., & De Baets, B. (2017). Distance metric learning : a two-phase approach. Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2017), 123–128. ESANN.
- Chicago author-date
- Nguyen Cong, Bac, Carlos Morell, and Bernard De Baets. 2017. “Distance Metric Learning : A Two-Phase Approach.” In Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2017), 123–28. ESANN.
- Chicago author-date (all authors)
- Nguyen Cong, Bac, Carlos Morell, and Bernard De Baets. 2017. “Distance Metric Learning : A Two-Phase Approach.” In Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2017), 123–128. ESANN.
- Vancouver
- 1.Nguyen Cong B, Morell C, De Baets B. Distance metric learning : a two-phase approach. In: Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2017). ESANN; 2017. p. 123–8.
- IEEE
- [1]B. Nguyen Cong, C. Morell, and B. De Baets, “Distance metric learning : a two-phase approach,” in Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2017), Bruges, Belgium, 2017, pp. 123–128.
@inproceedings{8532218,
abstract = {{Distance metric learning has been successfully incorporated in many machine learning applications. The main challenge arises from the positive semidefiniteness constraint on the Mahalanobis matrix, which results in a high computational cost. In this paper, we develop a novel approach to reduce this computational burden. We first map each training example into a new space by an orthonormal transformation. Then, in the transformed space, we simply learn a diagonal matrix. This two-phase approach is thus much easier and less costly than learning a full Mahalanobis matrix in one phase as is commonly done.}},
author = {{Nguyen Cong, Bac and Morell, Carlos and De Baets, Bernard}},
booktitle = {{Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2017)}},
isbn = {{9782875870391}},
language = {{eng}},
location = {{Bruges, Belgium}},
pages = {{123--128}},
publisher = {{ESANN}},
title = {{Distance metric learning : a two-phase approach}},
url = {{https://www.elen.ucl.ac.be/Proceedings/esann/esannpdf/es2017-8.pdf}},
year = {{2017}},
}