Regularized Common Spatial Pattern with Incomplete Generic Learning for EEG Classification in Small Sample Setting

Abstract:

Common spatial pattern (CSP) is based on sample covariance matrix estimation, whose estimation accuracy is severely overfit with small training sets. To address the drawback, the regularized CSP (R-CSP) was proposed that adds regularization information into the CSP learning process. In the algorithm, all samples of each generic subject were used for training sample covariance matrices. When only a part of the samples of each generic subject are allowed as generic training set, this R-CSP algorithm wouldn’t work. To solve this problem, an improved method is proposed in this paper. The new algorithm was applied to a brain-computer interface (BCI) data set containing five subjects and a mean improvement of 2.5% in classification rate was achieved.

• Periodical: Applied Mechanics and Materials (Volumes 303-306)
• Pages: 1344-1349
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.303-306.1344
• Online since: February 2013
• Authors: Hao Zhu, Qing Guo Wei
• Keywords: Brain Computer Interface (BCI), Common Spatial Pattern, Regularization, Small Sample Set

References

[1] J. R. Wolpaw, N. Birbaumer, D. J. McFarland, et al, Clinical Neurophysiol., 2002, 113: 767-91.

[2] A. Cichocki, Y. Washizawa, T. Rutkowski, et al, IEEE Trans. on Computer, 41(10): 34–42, (2008).

[3] T. Ebrahimi, J. F. Vesin, and G. Garcia, IEEE Signal Process. Mag., 2003, 20(1): 14–24.

[4] Z. J. Koles, Electroenc. Clin. Neurophys. 1991, 79: 440–447.

[5] Y. Li, X. Gao, and S. Gao, IEEE Trans. Biomed. Eng., 2004, 51(6): 1019–1025.

[6] H. Ramoser, J. Muller-Gerking, G. Pfurtscheller, IEEE Trans. Rehabil. Eng., 2000, 8(4): 441-6.

[7] G. Dornhege, B. Blankertz, G. Curio, et al, IEEE Trans. Biomed. Eng., 2004, 51( 6): 993-1002.

[8] J. Lu, K. N. Plataniotis, A. N. Venetsanopoulos, Pattern Recognit. Lett., 2003, 24(16): 3079-87.

[9] J. Lu, K. N. Plataniotis, A. N. Venetsanopoulos, Pattern Recognit. Lett., 2005, 26(2): 181–191.

[10] H. Lu, K. N. Plataniotis, A. N. Venetsanopoulos, Boosting LDA with regularization on MPCA features for gait recognition, in Proc. Biometrics Symposium 2007, September (2007).

DOI: 10.1109/bcc.2007.4430542

[11] H. Lu, K.N. Plataniotis, et al, IEEE Trans. Neural Networks, 2009, 20(1): 103–123.

[12] J. H. Friedman, Journal of the American Statistical Association, 1989, 84(405): 165–175.

[13] J. Farquhar, N.J. Hill, T.N. Lal and B. Schölkopf, Regularised CSP for Sensor Selection in BCI, in Proc. of the 3rd International BCI Workshop and Training Course, (2006).

[14] H. Lu, K. N. Plataniotis, and A. N. Venetsanopoulos, Regularized common spatial patterns with generic learning for EEG signal classification, in Proc. Eng. Med. Biol. Soc., Sep. (2009).

DOI: 10.1109/iembs.2009.5332554