Robust and sparse canonical correlation analysis based L 2,p -norm

Zhong-rong Shi; Sheng Wang; Chuan-cai Liu

Robust and sparse canonical correlation analysis based L _2,p-norm

View Fulltext

Author(s): Zhong-rong Shi¹ ; Sheng Wang² ; Chuan-cai Liu¹
- Affiliations: 1: School of Computer and Engineering , Nanjing University of Science and Technology , Nanjing 210094 , People's Republic of China ;
  2: Institute of Image Processing and Pattern Recognition, Henan University , Kaifeng 475004 , People's Republic of China
Source: Volume 2017, Issue 3, March 2017, p. 57 – 59
DOI: 10.1049/joe.2016.0296 , Online ISSN 2051-3305

This is an open access article published by the IET under the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0/)

Received 14/10/2016, Accepted 18/01/2017, Revised 23/12/2016, Published 30/03/2017

The objective function of canonical correlation analysis (CCA) is equivalent to minimising an L ₂-norm distance of the paired data. Owing to the characteristic of L ₂-norm, CCA is highly sensitive to noise and irrelevant features. To alleviate such problem, this study incorporates robust feature extraction and group sparse feature selection into the framework of CCA, and proposes a feature fusion method named robust and sparse CCA (RSCCA). In RSCCA, L _2,p-norm is adopted as the distance measurement of paired data, which can alleviate the effect of noise and irrelevant features and achieve robust performance. The experimental results show that our method outperforms CCA and its variants for feature fusion.

References

1. 1)
  - 21. Kwak, N.: ‘Principle component analysis based on L1-norm maximization’, IEEE Trans. Pattern Anal. Mach. Intell., 2008, 30, (9), pp. 1672–1680 (doi: 10.1109/TPAMI.2008.114).
2. 2)
  - 8. Wang, H., Nie, F., Huang, H.: ‘Robust and discriminative distance for multi-instance learning’. 2012 IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2012, pp. 2919–2924.
3. 3)
  - 4. Yuan, Y.H., Sun, Q.S., Zhou, Q., et al: ‘A novel multiset integrated canonical correlation analysis framework and its application in feature fusion’, Pattern Recognit., 2011, 44, (5), pp. 1031–1040 (doi: 10.1016/j.patcog.2010.11.004).
4. 4)
  - 16. Cai, D., He, X., Han, J., et al: ‘Orthogonal Laplacian faces for face recognition’, IEEE Trans. Image Process., 2006, 15, (11), pp. 3608–3614 (doi: 10.1109/TIP.2006.881945).
5. 5)
  - 6. Nie, F., Huang, H., Cai, X., et al: ‘Efficient and robust feature selection via joint L2, 1-norms minimization’, Advances in Neural Information Processing Systems 23 (NIPS 2010), Vancouver, Canada, December2010, pp. 1813–1821.
6. 6)
  - 2. De la Torre, F.: ‘A least-squares framework for component analysis’, IEEE Trans. Pattern Anal. Mach. Intell., 2012, 34, (6), pp. 1041–1055 (doi: 10.1109/TPAMI.2011.184).
7. 7)
  - 37. Yuan, Y.-H., Sun, Q.-S., Ge, H.-W.: ‘Fractional-order embedding canonical correlation analysis and its applications to multi-view dimensionality reduction and recognition’, Pattern Recognit., 2014, 47, pp. 1411–1424 (doi: 10.1016/j.patcog.2013.09.009).
8. 8)
  - 7. Peng, H., Fan, Y.: ‘Direct l(2, p)-norm learning for feature selection’, arXiv preprint arXiv:1504.00430, 2015.
9. 9)
  - 11. Sun, L., Ji, S., Ye, J.: ‘A least squares formulation for canonical correlation analysis’. Proc. of the 25th Int. Conf. on Machine Learning, 2008, pp. 1024–1031.
10. 10)
  - 1. Melzer, T., Reiter, M., Bischof, H.: ‘Appearance models based on kernel canonical correlation analysis’, Pattern Recognit.., 2003, 36, (9), pp. 1961–1971 (doi: 10.1016/S0031-3203(03)00058-X).
11. 11)
  - 10. Hou, S., Sun, Q.: ‘An orthogonal regularized CCA learning algorithm for feature fusion’, J. Vis. Commun. Image Represent., 2014, 25, (5), pp. 785–792 (doi: 10.1016/j.jvcir.2014.01.009).

Login

Not registered yet?

Share

Tools

Login to add to favourites

Key

Robust and sparse canonical correlation analysis based L _2,p-norm

References

Related content