Abstract
Both labeled and unlabeled data have been widely used in electroencephalographic (EEG)-based brain-computer interface (BCI). However, labeled EEG samples are generally scarce and expensive to collect, while unlabeled samples are considered to be abundant in real applications. Although the semi-supervised learning (SSL) allows us to utilize both labeled and unlabeled data to improve the classification performance as against supervised algorithms, it has been reported that unlabeled data occasionally undermine the performance of SSL in some cases. To overcome this challenge, we propose a collaborative representation-based semi-supervised extreme learning machine (CR-SSELM) algorithm to evaluate the risk of unlabeled samples by a new safety-control mechanism. Specifically, the ELM model is firstly used to predict unlabeled samples and then the collaborative representation (CR) approach is employed to reconstruct the unlabeled samples according to the obtained prediction results, from which the risk degree of unlabeled sample is defined. A risk-based regularization term is then constructed accordingly and embedded into the objective function of the SS-ELM. Experiments conducted on benchmark and EEG datasets demonstrate that the proposed method outperforms the ELM and SS-ELM algorithm. Moreover, the proposed CR-SSELM even offers the best performance while SS-ELM yields worse performance compared with its supervised counterpart (ELM).
Similar content being viewed by others
References
Ang KK, Guan C et al (2017) EEG-based strategies to detect motor imagery for control and rehabilitation. IEEE Tran Neural Syst Rehabil Eng 25(4):392–401
Pfurtscheller G, Neuper C et al (2001) Motor imagery and direct brain-computer communication. Proc IEEE 89(7):1123–1134
Lotte F, Bougrain L, Cichocki A, Clerc M, Congedo M, Rakotomamonjy A, Yger F (2018) A review of classification algorithms for eeg-based brain-computer interfaces: a 10-year update. J Neural Eng 15:031005
Ma YL, She QS et al (2016) Classification of motor imagery eeg signals with support vector machines and particle swarm optimization. Comput Math Methods Med 4941235
Li RH et al (2017) Enhancing performance of a hybrid EEG-fNIRS system using channel selection and early temporal features. Front Hum Neurosci 11:462
She QS, Gan HT, Ma YL, Luo ZZ (2016) Scale-dependent signal identification in low-dimensional subspace: motor imagery task classification. Neural Plast 7431012
Huang GB et al (2012) Extreme learning machine for regression and multiclass classification. IEEE Trans Syst Man Cybern B (Cybernetics) 42(2):513–529
Huang GB, Zhu QY, Siew CK (2006) Extreme learning machine: theory and applications. Neurocomputing 70(1–3):489–501
Liang NY, Saratchandran P et al (2006) Classification of mental tasks from eeg signals using extreme learning machine. Int J Neural Syst 16(01):29–38
Schwarz A, Brandstetter J, Pereira J, Müller-Putz GR (2019) Direct comparison of supervised and semi-supervised retraining approaches for co-adaptive BCIs. Med Biol Eng Comput 57(11):2347–2357
Gong C, Tao D, Maybank SJ, Liu W, Kang G, Yang J (2016) Multi-modal curriculum learning for semi-supervised image classification. IEEE Trans Image Process 25(7):3249–3260
Gao Y, Ma J, Yuille AL (2017) Semi-supervised sparse representation based classification for face recognition with insufficient labeled samples. IEEE Trans Image Process 26(5):2545–2560
Wang Z, Yoon S, Xie SJ, Lu Y, Park DS (2016) Visual tracking with semi-supervised online weighted multiple instance learning. Vis Comput 32(3):307–320
Belkin M, Niyogi P et al (2006) Manifold regularization: a geometric framework for learning from labeled and unlabeled examples. J Mach Learn Res 7(1):2399–2434
Huang G, Song S et al (2014) Semi-supervised and unsupervised extreme learning machines. IEEE Trans Cybern 44(12):2405–2417
Tu WT, Sun SL et al (2011) Semi-supervised feature extraction with local temporal regularization for EEG classification. International Joint Conference on Neural Networks, 6033202
Yang T, Priebe CE et al (2011) The effect of model misspecification on semi-supervised classification. IEEE Trans Pattern Anal Mach Intell 33(10):2093–2103
Gan H, Huang R, Luo Z, Fan Y, Gao F (2016) Towards a probabilistic semi-supervised kernel minimum squared error algorithm. Neurocomputing 171:149–155
She Q, Hu B, Gan H, Fan Y, Nguyen T, Potter T, Zhang Y (2018) Safe semi-supervised extreme learning machine for eeg signal classification. IEEE Access 6:49399–49407
Gan H, Li Z, Wu W, Luo Z, Huang R (2018) Safety-aware graph-based semi-supervised learning. Expert Syst Appl 107:243–254
Li YF, Zhou ZH et al (2011) Improving semi-supervised support vector machines through unlabeled instances selection. Proceedings of the 25th AAAI Conference on Artificial Intelligence, San Francisco, California, USA, 386-391
Li YF, Zhou ZH et al (2015) Towards making unlabeled data never hurt. IEEE Trans Pattern Anal Mach Intell 37:175–188
Wang Y, Chen S et al (2013) Safety-aware semi-supervised classification. IEEE Trans Neural Netw Learn Syst 24(11):1763–1772
Gan H, Luo ZZ, Meng M, Ma Y, She Q (2016) A risk degree-based safe semi-supervised learning algorithm. Int J Mach Learn Cybern 7(1):85–94
Gan H, Luo ZZ et al (2016) Towards designing risk-based safe laplacian regularized least squares. Expert Syst Appl 45:1–7
Gan H, Li Z, Fan Y, Luo Z (2018) Dual learning-based safe semi-supervised learning. IEEE Access 6:2615–2621
Xia Y, He D et al (2016) Dual learning for machine translation. Adv Neural Inf Proces Syst 29:820–828
Dua D, Graff C (2019) UCI machine learning repository. Irvine, CA: University of California, School of Information and Computer Science. [Online]. Available: http://archive.ics.uci.edu/ml
Schroff F, Treibitz T, Christiansen E (2013) Yale face database. University of California, san Diego, USA. [online]. Available: http://vision.ucsd.edu/content/yale-face-database
Gan H, Sang N, Huang R (2015) Manifold regularized semi-supervised gaussian mixture model. J Opt Soc Am A 32(4):566–575
Michael T, Müller Klaus-Robert Ad A et al (2012) Review of the BCI competition IV. Front Neurosci 00035
Keng AK, Yang CZ et al (2012) Filter bank common spatial pattern algorithm on BCI competition IV datasets 2a and 2b. Front Neurosci 00039
Meng M, Zhu JQ et al (2016) Two-level feature extraction method for multi-class motor imagery EEG. Acta Automat Sin 42:1915–1922
Chiu SC, Jin Z et al (2015) Active learning combining uncertainty and diversity for multi-class image classification. IET Comput Vis 9(3):400–407
Deng J, Zhang Z et al (2013) Sparse autoencoder-based feature transfer learning for speech emotion recognition. Proceedings of the 5th Biannual Humaine Association Conference on Affective Computing and Intelligent Interaction, 511-516
Peng Y, Wang S, Long X, Lu BL (2015) Discriminative graph regularized extreme learning machine and its application to face recognition. Neurocomputing 149:340–353
Kong Y, Wang XS et al (2018) Hyperspectral imagery classification based on semi-supervised broad learning system. Remote Sens 10(5):685
Acknowledgments
The authors would like to acknowledge the BCI Competition IV Dataset 2a which is used to test the algorithms proposed in this study.
Funding
This work is supported by National Nature Science Foundation under Grant (No. 61871427).
Author information
Authors and Affiliations
Contributions
Conceptualization, Qingshan She; methodology, Qingshan She; software, Jie Zou; validation, Qingshan She and Jie Zou; formal analysis, Zhizeng Luo; investigation, Thinh Nguyen; resources, Rihui Li; writing—original draft preparation, Qingshan She; writing—review and editing, Qingshan She and Jie Zou; supervision, Yingchun Zhang.
Corresponding authors
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
She, Q., Zou, J., Luo, Z. et al. Multi-class motor imagery EEG classification using collaborative representation-based semi-supervised extreme learning machine. Med Biol Eng Comput 58, 2119–2130 (2020). https://doi.org/10.1007/s11517-020-02227-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11517-020-02227-4