基于混合迁移学习的运动想象分类算法研究及其在脑机接口中的应用

doi:10.3969/j.issn.1000-1158.2021.05.14

摘要
图/表
参考文献(27)
相关文章 (15)

全文: PDF (2639 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要为提升迁移学习在运动想象脑机接口应用过程中的迁移高效性及普适性,综合实例迁移和特征迁移学习方法的优势进而构建了混合迁移学习模型。首先,依据样本权重极化原理改进TrAdaBoost算法以实现实例层面的迁移,优化源域训练样本;其次,基于大间隔投射迁移支持向量机进一步缩短源域与目标域间的分布距离以完成特征层面的迁移,实现迁移效率最大化。进一步,将该方法应用于脑机接口竞赛Dataset IIb数据集进行离线测试及分析,研究结果表明混合迁移学习模型的迁移效率明显高于单一迁移学习模型,并且对于不同迁移对象识别准确率相对提升均值在70%以上,验证了所述方法的有效性与普适性。此外,基于已搭建的运动想象识别系统进行在线测试,验证了模型的实用性。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	杜义浩
	刘兆军
	付子豪
	张园园
	任娜
	陈杰
	谢平

关键词 ：计量学, 运动想象, 脑机接口, 实例迁移, 特征迁移, 混合迁移

Abstract：To improve the efficiency and universality of transfer learning in the application of motor imagery brain-computer interface (MI-BCI),a hybrid transfer learning model with integrating the advantages of instance transfer and feature transfer learning methods was built.Firstly,the transfer of the instance level by introducing the principle of sample weight polarization to improve the classical TrAdaBoost algorithm was realized,which can optimize training samples in the source domain to some extent.Secondly,to further narrow the distance between the source domain and the target domain,the large margin projected transductive support vector machine was applied to complete the transfer of the feature level,thus maximizing the transfer efficiency.Furthermore,the proposed method was applied to the BCI competition dataset (Dataset IIb data set) for offline test and analysis.The results showed that the hybrid transfer learning model achieved significantly better transfer efficiency than the single transfer learning model,and obtained an improvement of the recognition rate with average value of above 70% for different transfer objects.The results also verified the effectiveness and universality of the hybrid transfer learning model.In addition,the online test was carried out based on the established motor-imagery system,which further verified the practicability of the model.

Key words： metrology motor imagery brain-computer interface instance transfer feature transfer hybrid transfer

收稿日期: 2019-09-27 发布日期: 2021-05-24

PACS:	TB973
	TB99

基金资助:国家自然科学基金(61673336);河北省自然科学基金(F2015203372);河北省高等学校科学技术研究项目(QN2016094)

通讯作者: 谢平(1972-),黑龙江齐齐哈尔人,燕山大学教授,主要从事脑机接口、智能康复等方面的研究。Email:pingx@ysu.edu.cn E-mail: pingx@ysu.edu.cn

作者简介: 杜义浩(1983-), 河北沧州人,燕山大学讲师,主要从事康复机器人生物反馈控制等方面的研究。Email: duyihao@126.com

引用本文:

杜义浩,刘兆军,付子豪,张园园,任娜,陈杰,谢平. 基于混合迁移学习的运动想象分类算法研究及其在脑机接口中的应用[J]. 计量学报, 2021, 42(5): 629-637.
DU Yi-hao,LIU Zhao-jun,FU Zi-hao,ZHANG Yuan-yuan,REN Na,CHEN Jie,XIE Ping. Motion Imagery Classification Algorithm Research Based on Hybrid Transfer Learning and Application in Brain-computer Interface. Acta Metrologica Sinica, 2021, 42(5): 629-637.

链接本文:

http://jlxb.china-csm.org:81/Jwk_jlxb/CN/10.3969/j.issn.1000-1158.2021.05.14 或 http://jlxb.china-csm.org:81/Jwk_jlxb/CN/Y2021/V42/I5/629

［1］高上凯. 浅谈脑-机接口的发展现状与挑战［J］. 中国生物医学工程学报, 2007, 26 (6): 801-803.
Gao S K. Comments on Recent Progress and Challenges in the Study of Brain-Computer Interface［J］. Chinese Journal of Biomedical Engineering, 2007, 26 (6): 801-803.
［2］张丹, 陈菁菁, 王毅军. 2017年脑机接口研发热点回眸［J］. 科技导报, 2018, 36 (1): 104-109.
Zhang D, Chen J J, Wang Y J. An overeview of the forntier on braincomputer interface technology in 2017［J］. Science & Technology Review, 2018, 36 (1): 104-109.
［3］Millán J D R, Rupp R, Müller-Putz G R, et al. Com-bining Brain-Computer Interfaces and Assistive Techn-ologies: State-of-the-Art and Challenges［J］. Frontiers in Neuroscience, 2010, 4 (5): 161.
［4］Cao X D, Wipf D, Wen F, et al. A Practical Transfer Learning Algorithm for Face Verification［C］//IEEE Computer Society. 2013 IEEE International Conference on Computer Vision (IC-CV). Sydney, NSW, 2013.
［5］胡学钢, 方玉成, 张玉红. 基于Logistic回归分析的直推式迁移学习［J］. 合肥工业大学学报 (自然科学版), 2010, 33 (12): 1797-1801.
Hu X G, Fang Y C, Zhang Y H. Logistic regression for transductive transfer learning［J］. Journal of Hefei Uni-versity of Technology (Natural Science), 2010, 33 (12): 1797-1801.
［6］刘三民, 刘余霞. 基于实例迁移的数据流分类挖掘方法［J］. 信息与控制, 2019, 48 (3): 380-384.
Liu S M, Liu Y X. Classification Mining Method for Data Streams Based on Instances Transfer［J］. Information and Control, 2019, 48 (3): 380-384.
［7］阎高伟, 贺敏, 汤健, 等. 基于最大均值差异多源域迁移学习的湿式球磨机负荷参数软测量［J］. 控制与决策, 2018, 33 (10): 70-75.
Yan G W, He M, Tang J, et al. Soft sensor of wet ball mill load based on maximum mean discrepancy multi-source domain transfer learning［J］. Control and Dec-ision, 2018, 33 (10): 70-75.
［8］韩飞. 基于运动想象的脑电特征提取及特征迁移方法研究［D］. 哈尔滨: 哈尔滨工业大学, 2017.
［9］Pan S J, Kwok J T, Yang Q. Transfer Learning via Dimensionality Reduction［C］// Proceedings of the Twenty-Third AAAI Conference on Artificial Intelligence, AAAI 2008, Chicago, USA, 2008.
［10］Dai W, Yang Q, Xue G R, et al. Boosting for transfer learning［C］// International Conference on Machine Learning, Oregon, USA, 2007.
［11］林程. 基于迁移学习的脑机接口信号分类算法研究［D］. 广州: 华南理工大学, 2017.
［12］Quanz B, Huan J. Large margin transductive transfer learning［C］// Proceedings of the 18th ACM Conference on Information and Knowledge Management, Hong Kong, China, 2009.
［13］何群, 杜硕, 王煜文, 等. 基于变分模态分解与深度信念网络的运动想象分类识别研究［J］. 计量学报, 2020, 41 (1): 90-99.
He Q, Du S, Wang Y W, et al. The Classification of EEG Induced by Motor Imagery Based on Variational Mode Decomposition and Deep Belief Network［J］. Acta Metrologica Sinica, 2020, 41 (1): 90-99.
［14］Blankertz B, Klaus-Robert Müller, Krusienski D J, et al. The BCI competition III: Validating alternative appr-oachs to actual BCI problems［J］. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2006, 14 (2): 153-159.
［15］Pfurtscheller G. Functional brain imaging based on ERD/ERS［J］. Vision research, 2001, 41 (10-11): 1257.
［16］Pfurtscheller G, Lopes de Silva F H. Event-related EEG/MEG synchronization and desynchronization: basic principles［J］. Clinical Neurophysiology, 1999, 110 (11): 1842-1857.
［17］Hjorth B. EEG analysis based on time domain properties［J］. Electroencephalography & Clinical Neurophysiology, 1970, 29 (3): 0-310.
［18］Oh S H, Lee Y R. A novel eeg feature extraction method using hjorth parameter［J］. International Journal of Electronics and Electrical Engineering, 2014, 2 (2): 106-110.
［19］伍广腾. 基于运动想象的脑-机接口在线异步系统研究［D］. 哈尔滨: 哈尔滨工业大学, 2018.
［20］刘琳琳, 陈健, 李松, 等. 基于相位同步与AR的运动想象脑电信号特征提取研究［J］. 软件导刊, 2018, 17 (3): 7-10.
Liu L L, Chen J, Li S, et al. Feature Extraction of EEG in Motion Imagery Based on Phase Synchronization and AR［J］. Software Guide, 2018, 17 (3): 7-10.
［21］胡春海, 李涛, 刘永红, 等. 基于改进差分进化算法的运动想象脑机接口频带选择［J］. 计量学报, 2018, 39 (2): 134-137.
Hu C H, Li T, Liu Y H, et al. Motor Imagrey Brain Computer Interface Band Selection Based on Improved Differential Evolution Algorithm［J］. Acta Metrologica Sinica, 2018, 39 (2): 276-279.
［22］Gong P, Chen M Y, Zhang L, et al. HHT-Based Selection of Optimal Time-Frequency Patterns for Motor Imagery［J］. Applied Mechanics & Materials, 2012, 380-384 (22): 3522-3525.
［23］谢平, 陈晓玲, 苏玉萍, 等. 基于EMD-多尺度熵和ELM的运动想象脑电特征提取和模式识别［J］. 中国生物医学工程学报, 2013, 32 (6): 641-648.
Xie P, Chen X L, Su Y P, et al. Feature Extraction and Recognition of Motor Imagery EEG Based on EMD-Multiscale Entropy and Extreme Learning Machine［J］. Chinese Journal of Biomedical Engineering, 2013, 32 (6): 641-648.
［24］孟宗, 顾海燕, 刘利晖, 等. 基于EMD与AR谱的轧机主传动系统故障诊断研究［J］. 计量学报, 2011, 32 (4): 338-342.
Meng Z, Gu H Y, Liu L H, et al. Study on Fault Diagnosis of Main Drive System of Rolling Mill Based on EMD and AR Spectrum［J］. Acta Metrologica Sinica, 2011, 32 (4): 338-342.
［25］Dragomiretskiy K, Zosso D. Variational Mode Decompo-sition［J］. IEEE Transactions on Signal Processing, 2014, 62 (3): 531-544.
［26］胡春海, 信思旭, 刘斌, 等. 基于小波变换和盲源分离的P300识别算法研究［J］. 计量学报, 2017, 38 (2): 242-246.
Hu C H, Xin S X, Liu B, et al. Study on Recognition Algorithm of P300 Based on Wavelet Transform and Blind Source Separation［J］. Acta Metrologica Sinica, 2017, 38 (2): 242-246.
［27］李晓竹, 解华. 基于 AdaBoost 的多项式旋流器溢流粒度软测量模型［J］. 计量学报, 2016, 37 (3): 275-278.
Li X Z, Xie H. Polynomial soft sensor model of the overflow particle size of cyclone based on AdaBoost［J］. Acta Metrologica Sinica, 2016, 37 (3): 275-278.