Self-weighted Robust LDA for Multiclass Classification with Edge Classes

ACM Transactions on Intelligent Systems and Technology - Tập 12 Số 1 - Trang 1-19 - 2021
Caixia Yan1, Xiaojun Chang2, Minnan Luo1, Qinghua Zheng1, Xiaoqin Zhang3, Zhihui Li4, Feiping Nie5
1School of Electronic and Information Engineering, Xi’an Jiaotong University, Shaanxi, China
2Faculty of Information Technology, Monash University, Australia
3College of Computer Science and Artificial Intelligence, Wenzhou University, China
4Qilu University of Technology (Shandong Academy of Sciences), Jinan, Shandong, China
5Center for Optical Image Analysis and Learning, Northwestern Polytechnical University, Shaanxi, China

Tóm tắt

Linear discriminant analysis (LDA) is a popular technique to learn the most discriminative features for multi-class classification. A vast majority of existing LDA algorithms are prone to be dominated by the class with very large deviation from the others, i.e., edge class, which occurs frequently in multi-class classification. First, the existence of edge classes often makes the total mean biased in the calculation of between-class scatter matrix. Second, the exploitation of ℓ2-norm based between-class distance criterion magnifies the extremely large distance corresponding to edge class. In this regard, a novel self-weighted robust LDA with ℓ2,1-norm based pairwise between-class distance criterion, called SWRLDA, is proposed for multi-class classification especially with edge classes. SWRLDA can automatically avoid the optimal mean calculation and simultaneously learn adaptive weights for each class pair without setting any additional parameter. An efficient re-weighted algorithm is exploited to derive the global optimum of the challenging ℓ2,1-norm maximization problem. The proposed SWRLDA is easy to implement and converges fast in practice. Extensive experiments demonstrate that SWRLDA performs favorably against other compared methods on both synthetic and real-world datasets while presenting superior computational efficiency in comparison with other techniques.

Từ khóa


Tài liệu tham khảo

Abou-Moustafa Karim T., Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.

Bian Wei, 2008, Proceedings of the International Conference on Pattern Recognition.

10.1109/TPAMI.2010.189

C. L. Blake and C. J. Merz. 1998. UCI Repository of Machine Learning Databases. Department of Information and Computer Science University of California Irvine CA. C. L. Blake and C. J. Merz. 1998. UCI Repository of Machine Learning Databases. Department of Information and Computer Science University of California Irvine CA.

10.1109/TKDE.2007.190669

10.1145/1961189.1961199

Chang Xiaojun, Proceedings of the European Conference on Computer Vision, Vittorio Ferrari, Martial Hebert, Cristian Sminchisescu, and Yair Weiss (Eds.).

Chang Xiaojun, 2015, Proceedings of the AAAI Conference on Artificial Intelligence, Blai Bonet and Sven Koenig (Eds.).

Xiaojun Chang Haoquan Shen Sen Wang Jiajun Liu and Xue Li. 2014. Semi-supervised feature analysis for multimedia annotation by mining label correlation. In Proceedings of the 18th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining (Lecture Notes in Computer Science) Vincent S. Tseng Tu Bao Ho Zhi-Hua Zhou Arbee L. P. Chen and Hung-Yu Kao (Eds.) Vol. 8444. Springer 74--85. Xiaojun Chang Haoquan Shen Sen Wang Jiajun Liu and Xue Li. 2014. Semi-supervised feature analysis for multimedia annotation by mining label correlation. In Proceedings of the 18th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining (Lecture Notes in Computer Science) Vincent S. Tseng Tu Bao Ho Zhi-Hua Zhou Arbee L. P. Chen and Hung-Yu Kao (Eds.) Vol. 8444. Springer 74--85.

10.1109/TNNLS.2016.2582746

Chang Xiaojun, Proceedings of the Conference on Computer Vision and Pattern Recognition.

10.24963/ijcai.2017/134

10.1016/j.patcog.2018.05.007

10.1016/0375-6742(92)90105-H

10.5555/1059467.1059472

Fanty Mark, 1991, Proceedings of the Conference on Neural Information Processing Systems.

10.1016/j.eswa.2013.08.089

Fukunaga Keinosuke, Introduction to Statistical Pattern Recognition

10.1109/34.927464

10.1109/TCYB.2017.2773562

10.1109/CVPR.2018.00946

10.1006/jmva.1999.1857

10.1007/3-540-36755-1_14

10.1016/S0167-9473(02)00299-2

10.1109/TNNLS.2013.2258937

10.5555/3091622.3091662

10.1109/TNNLS.2018.2817538

10.1109/TKDE.2017.2728531

10.1109/ACCESS.2018.2859299

10.1109/34.935849

10.1109/TCYB.2017.2647904

Luo Minnan, 2016, Proceedings of the International Joint Conference on Artificial Intelligence, Subbarao Kambhampati (Ed.).

10.1109/TNNLS.2017.2650978

Douglas Martin R., 2006, Yohai

10.1109/TCYB.2016.2577590

Peng Chong, 2016, A supervised learning model for high-dimensional and large-scale data, ACM Trans. Intell. Syst. Technol., 8, 1, 10.1145/2972957

Peng Chong, 2019, Discriminative ridge machine: A classifier for high-dimensional data or imbalanced data. arXiv preprint arXiv:1904.07496

Rao C. Radhakrishna, 1948, The utilization of multiple measurements in problems of biological classification, J. Royal Stat. Soc., 10, 159, 10.1111/j.2517-6161.1948.tb00008.x

10.5555/874061.875452

Gilbert W. Stewart. 2001. Eigensystem. Society for Industrial 8 Applied Mathematics. Gilbert W. Stewart. 2001. Eigensystem. Society for Industrial 8 Applied Mathematics.

10.1109/TPAMI.2008.70

10.1007/s13042-016-0540-0

10.1016/j.sigpro.2014.12.012

10.1109/TCSVT.2018.2799214

10.1145/2956556

10.1145/3007195

10.1109/ICPR.2010.144

Xue Xiaowei, 2017, In Proceedings of the AAAI Conference on Artificial Intelligence, Satinder P. Singh and Shaul Markovitch (Eds.).

Syed Yahaya Sharipah Soaad, 2016, Robust linear discriminant analysis with automatic trimmed mean, J. Telecommun. Electron. Comput. Eng., 8, 1

10.1016/j.ins.2017.08.004

10.1145/1273496.1273633

10.1109/TKDE.2006.160

10.1145/1143844.1143979

10.1016/j.patcog.2010.09.019

10.1109/TIP.2017.2685343

Zhang Liang, Proceedings of the International Conference on Acoustics, Speech and Signal Processing.

10.1007/978-3-540-88682-2_55

10.1109/TNNLS.2020.3009209

10.1109/ICFHR.2012.196

Zhang Yu, 2010, Proceedings of the Conference on Neural Information Processing Systems.

10.1162/neco_a_01116

10.1109/TKDE.2018.2842023

10.1162/NECO_a_00950

10.1109/TCYB.2016.2529299

Zhu Lei, Proceedings of the ACM International Conference on Multimedia, Qiong Liu, Rainer Lienhart, Haohong Wang, Sheng-Wei “Kuan-Ta