An iterative self-training support vector machine algorithm in brain-computer interfaces

Abstract

In this paper, an iterative self-training Support Vector Machine (SVM) algorithm combined feature re-extraction is proposed for semi-supervised learning, which only needs a small set of labeled samples to train classifier and is thus very useful in Brain-Computer Interface (BCI) design. Two methods, the model selection based self-training and the confidence criterion, respectively, is also proposed for searching the best parameter pair of SVM and selecting the most useful unlabeled data to expand the labeled training data set. The Dataset IVa of BCI Competition III, is presented to demonstrate the validity of our algorithm with statistical significance test. As an iterative algorithm, experimental results of the proposed algorithm show the validity of re-extracting feature and the robustness of the feature to the noise. In addition, the convergence of the proposed algorithm and the validity of the method measuring the consistency of the feature are also demonstrated in experiments.

Keywords

Self-training support vector machine (SVM)brain-computer interface (BCI)model selection confidence criterion

References

Arvaneh

, Guan

, Ang

K.K.

and Quek

, EEG data space adaptation to reduce intersession nonstationarity in brain-computer interface, in: Neural Computation 25(8) (2013), 2146-2171.

Vidaurre

, Kawanabe

, von Bünau

, Blankertz

and Müller

K.R.

, Toward unsupervised adaptation of LDA for brain-computer interfaces, IEEE Transactions on Biomedical Engineering 58(3) (2011), 587-597.

Keyvanpour

M.R.

and Mani

M.B.

, Semi-supervised text categorization: Exploiting unlabeled data using ensemble learning algorithms, in: Intelligent Data Analysis 17(3) (2013), 367-385.

Panicker

R.C.

, Puthusserypady

and Sun

, daptation in P300 brain-computer interfaces: A two-classifier cotraining approach, in: IEEE Transactions on Biomedical Engineering 57(12) (2010), 2927-2935.

Liao

, Yao

and Li

, Transductive SVM for reducing the training effort in BCI, in: Journal of Neural Engineering 4(3) (2007), 246.

Liu

, Jun

and Ghosh

, self-training approach to cost sensitive uncertainty sampling, in: Machine Learning 76(2) (2009), 257-270.

Diplaros

, Vlassis

and Gevers

, A spatially constrained generative model and an EM algorithm for image segmentation, in: IEEE Transactions on Neural Networks 18(3) (2007), 798-808.

Nigam

, McCallum

and Mitchell

, Semi-supervised text classification using EM, in: Semi-supervised Learning, Zien

, ed., Cambridge, MA: MIT Press, 2006, pp. 33-56.

Ramoser

, Muller-Gerking

and Pfurtscheller

, Optimal spatial filtering of single trial EEG during imagined hand movement, in: IEEE Transactions on Rehabilitation Engineering 8(4) (2000), 441-446.

10.

Reuderink

and Poel

, Robustness of the common spatial patterns algorithm in the BCI-pipeline, in: Technical Report, HMI, University of Twente (2008).

11.

Ang

K.K.

, Zhang

Z.Y.

, Zhang

and Guan

, Filter bank common spatial pattern (FBCSP) in brain-computer interface, in: Proceedings of the IEEE International Joint Conference on Computational Intelligence (2008), 2390-2397.

12.

Ang

K.K.

, Chin

Z.Y.

, Wang

, Guan

and Zhang

, Filter bank common spatial pattern algorithm on BCI competition iv datasets 2a and 2b, in: Frontiers in Neuroscience 6(39) (2012).

13.

Zhou

, Bousquet

, Lal

T.N.

, Weston

and Scholkopf

, Learning with local and global consistency, in: NIPS 16 (2003), 321-328.

14.

and Guan

, An extended EM algorithm for joint feature extraction and classification in brain-computer interfaces, in: Neural Computation 18(11) (2006), 2730-2761.

15.

and Guan

, Joint feature re-extraction and classification using an iterative semi-supervised support vector machine algorithm, in: Machine Learning 71(1) (2008), 33-53.

16.

Bamdadian

, Guan

, Ang

K.K.

et al., Online semi-supervised learning with KL distance weighting for motor imagery-based BCI, in: Proceedings of the IEEE Annual International Conference on Medicine and Biology Society (EMBC) (2012), 2732-2735.

17.

Ang

K.K.

, Chin

Z.Y.

, Zhang

and Guan

, Filter bank common spatial pattern (FBCSP) algorithm using online adaptive and semi-supervised learning, in: Proc of the 2011 International Joint Conference on Neural Networks (2011), 392-396.

18.

Ang

K.K.

, Chin

Z.Y.

, Zhang

and Guan

, Mutual information-based selection of optimal spatial-temporal patterns for single-trial EEG-based BCIs, in: Pattern Recognition 45(6) (2012), 2137-2144.

19.

Campbell

, Cristianini

and Smola

, Query learning with large margin classifiers, in: Proceedings of the 2000 ICML (2000), 111-118.

20.

, Guan

, Li

et al., A self-training semi-supervised SVM algorithm and its application in an EEG-based brain computer interface speller system, in: Pattern Recongnition Letters 29(9) (2008), 1285-1294.

21.

Chang

C.C.

and Lin

C.J.

, LIBSVM: A library for support vector machines, in: ACM Transactions on Intelligent Systems and Technology (TIST) 2(3) (2011), 27.