[1]庄玮,段锁林,徐亭婷,等.基于SVM的4类运动想象的脑电信号分类方法[J].常州大学学报(自然科学版),2014,(01):42-46.[doi:10.3969/j.issn.2095-0411.2014.01.010]
　ZHUANG Wei,DUAN Suo-lin,XU Ting-ting.Research on Classification Method Based on SVM for the FourClass Motor Imagery EEG[J].Journal of Changzhou University(Natural Science Edition),2014,(01):42-46.[doi:10.3969/j.issn.2095-0411.2014.01.010]
点击复制

基于SVM的4类运动想象的脑电信号分类方法()
分享到：

常州大学学报(自然科学版)[ISSN:2095-0411/CN:32-1822/N]

卷:
期数:
2014年01期
页码:
42-46
栏目:
出版日期:
2014-02-28

文章信息/Info

Title:
Research on Classification Method Based on SVM for the FourClass Motor Imagery EEG
作者:
庄玮；段锁林；徐亭婷；
常州大学 机器人研究所,江苏 常州 213164
Author(s):
ZHUANG Wei; DUAN Suo-lin; XU Ting-ting
Institute of Robotics,Changzhou University,Changzhou 213164,China
关键词:
脑机接口;  4类运动想象;  特征提取;  聚类思想;  支持向量机
Keywords:
braincomputer interface(BCI);  fourclass motor imagery;  feature extraction;  clustering idea;  support sector machines(SVM)
分类号:
TP242.6
DOI:
10.3969/j.issn.2095-0411.2014.01.010
文献标志码:
A
摘要:
针对传统支持向量机分类方法在脑电信号处理中存在分类正确率低的问题,将聚类思想与二叉树支持向量机结合构造多类SVM分类器。实验以“BCI Competition 2005”中的Dataset Ⅲa为例,先对采集的4类运动想象脑电信号应用小波变换进行去噪; 再在分析小波包频带划分特点的基础上,利用小波包进行分解与重构,获取相应的能量特征; 最后应用改进后的支持向量机(SVM)分类方法对特征信号进行分类。结果表明该方法分类正确率较高,可以达到91.12%,并且有效的减少了分类器的个数,最终达到较好的识别效果。
Abstract:
For the disadvantages of the traditional SVM classificationin dealing with EEG signal,such as lower accuracy rate in classification,a multiclass SVM classifier is constructed by combining cluster idea with binary tree SVM.Based on data of the Dataset Ⅲa in the “BCI Competition 2005”.Firstly,fourclass motor imagery EEG data collected is denoised by the wavelet transform.Secondly,on the basis of analyzing the frequency band feature of wavelet packets,the corresponding energy feature is extracted by using decomposition and reconstruction of wavelet packets.Finally,the classification of the obtained feature signal is completed by using the improved SVM classification method.The simulation results show that the higher accuracy rate in the classification,about 91.12%,can be achieved.The number of classifier can be reduced efficiently and therelatively good identifying effects can be achieved finally.

参考文献/References:

[1]Millan del R J.On the need for on-line learning in brain-computer interfaces[C]∥Proceedings of 2004 IEEE International Joint Conference on Neural Networks.Budapest:Institute of Electrical and Electronics Engineers Incorporation Publisher,2004:2877-2882.
[2]WILSON J A,MELLINGER J,SCHALK G,et al.A procedure for measuring latencies in brain-computer interfaces[J].IEEE Transactions on Biomedical Engineering,2010,57(7):1785-1797.
[3]Vapnik V N.统计学习理论的本质[M].北京:清华大学出版社,2000.
[4]万柏坤,刘延刚,明东,等.基于脑电特征的多模式想象动作识别[J].天津大学学报:自然科学与工程技术版,2010,43(10):895-900.
[5]YANG B H,YANA G ZH,YAN R G,et a1.Adaptive subject-based feature extraction in brain-computer interfaces using wavelet packet best basis decomposition[J].Medical Engineering & Physics,2007,29(1):48-53.
[6]蔡新波.基于脑电信号的电动轮椅控制方法研究[D].杭州:杭州电子科技大学,2012.
[7]黄玲,张爱华.改进的决策树SVM在脑电识别中的应用[J].计算机工程与设计,2010,31(2):382-384.
[8]赵志刚,吕慧显,李玉景,等.一种基于聚类思想的SVM多类分类方法[J].青岛理工大学学报:自然科学版,2011,32(1):73-76.
[9]薛欣,贺国平.基于隶属度分离测度SVM决策树层次结构设计方法[J].计算机应用研究,2007,24(9):162-167.
[10]TANG Yan,TANG Jing-tian,GONG An-dong.Multi-class EEG classification forbrain computer interface based on CSP[C]∥Proceedings of the 1st International Conference on BioMedical Engineering and Informatics.Sanya:Institute of Electrical and Electronics Engineers Computer Society Press,2008:469-472.

相似文献/References:

[1]刘 成,何可人,周天彤,等.左右手运动想象脑电模式识别研究[J].常州大学学报(自然科学版),2013,(01):25.[doi:10.3969/j.issn.2095-0411.2013.01.005]
　LIU Cheng,HE Ke-ren,ZHOU Tian-tong,et al.Study of Discrimination between Left and Right Hand Movements Imagery Event-Related EEG Pattern[J].Journal of Changzhou University(Natural Science Edition),2013,(01):25.[doi:10.3969/j.issn.2095-0411.2013.01.005]

备注/Memo

备注/Memo:
基金项目:机器人技术与系统国家重点实验室开放基金重点项目(SKLRS20102D09) 作者简介:庄玮(1988- ),女,山东济南人,硕士生; 通讯联系人:段锁林。
更新日期/Last Update: 2014-02-20