Abstract
Linear discriminant analysis (LDA) is a popular supervised method for dimensionality reduction for its simplicity and effectiveness. However, its performance may significantly deteriorate in the situation that labeled training samples are very scarce and many semi-supervised methods are presented to solve the problem with a large amount of unlabeled samples available. In this paper, inspired by the uncertainty idea, we propose a semi-supervised uncertain linear discriminant analysis (SULDA). First, we present a fractional-step label propagation to reliably label the unlabeled samples. Secondly, we use all training samples and their labels to construct double-layer Gaussian mixture models. Finally, we utilize the originally labeled samples and the obtained models to define expected between-class and within-class scatter matrices. Experimental results illustrate that our method is superior to some state-of-the-art semi-supervised methods with respect to the discriminative power.
Supported by Hunan Provincial Natural Science Foundation of China (Grant No. 2018JJ2133).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Turk, M., Pentland, A.: Eigenfaces for recognition. J. Cogn. Neurosci. 3(1), 71–86 (1991)
Belhumeur, P.-N., Hespanha, J.-P., Kriegman, D.-J.: Eigenfaces vs. Fisherfaces: recognition using class specific linear projection. IEEE Trans. Pattern Anal. Mach. Intell. 19(7), 711–720 (1997)
Cai, D., He, X., Han, J.: Semi-supervised discriminant analysis. In: Proceedings of IEEE 11th International Conference on Computer Vision, pp. 1–7 (2007)
Huang, Y., Xu, D., Nie, F.: Semi-supervised dimension reduction using trace ratio criterion. IEEE Trans. Neural Netw. Learn. Syst. 23(3), 519–526 (2012)
Sindhwani, V., Niyogi, P., Belkin, M., Keerthi, S.: Linear manifold regularization for large scale semi-supervised learning. In: Proceedings of the 22nd ICML Workshop on Learning with Partially Classified Training Data, pp. 1–7 (2005)
Nie, F., Xu, D., Tsang, I.-W.-H., Zhang, C.: Flexible manifold embedding. IEEE Trans. Image Process. 19(7), 1921–1932 (2010)
Nie, F., Xiang, S., Jia, Y.: Semi-supervised orthogonal discriminant analysis via label propagation. Pattern Recogn. 42(11), 2615–2627 (2009)
Zhao, M., Zhang, Z., Chow, T.-W., Li, B.: A general soft label based linear discriminant analysis for semi-supervised dimensionality reduction. Neural Netw. 55, 83–97 (2014)
Wang, S., Lu, J., Gu, X., Du, H., Yang, J.: Semi-supervised linear discriminant analysis for dimension reduction and classification. Pattern Recogn. 57, 179–189 (2016)
Li, C.-N., Shao, Y.-H., Deng, N.-Y.: Robust L1-norm two-dimensional linear discriminant analysis. Neural Netw. 65, 92–104 (2015)
Wang, H., Lu, X., Hu, Z., Zheng, W.: Fisher discriminant analysis with L1-norm. IEEE Trans. Cybern. 44(6), 828–842 (2014)
Saeidi, R., Ramón, F.-A., Kolossa, D.: Uncertain LDA: including observation uncertainties in discriminative. IEEE Trans. Pattern Anal. Mach. Intell. 38(7), 1479–1488 (2016)
Bishop, C.-M.: Pattern Recognition and Machine Learning, 2nd edn. Springer, Heidelberg (2006)
Nene, S.-A., Nayar, S.-K., Murase, H.: Columbia object image library (coil-20). Columbia University, Department of Computer Science, New York (1996)
Hull, J.-J.: A database for handwritten text recognition research. IEEE Trans. Pattern Anal. Mach. Intell. 16(5), 550–554 (1999)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Shao, G., Liu, F., Peng, C. (2020). Semi-supervised Uncertain Linear Discriminant Analysis. In: Peng, Y., et al. Pattern Recognition and Computer Vision. PRCV 2020. Lecture Notes in Computer Science(), vol 12307. Springer, Cham. https://doi.org/10.1007/978-3-030-60636-7_13
Download citation
DOI: https://doi.org/10.1007/978-3-030-60636-7_13
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-60635-0
Online ISBN: 978-3-030-60636-7
eBook Packages: Computer ScienceComputer Science (R0)