华中科技大学学报（自然科学版）2011，Vol.39Issue(z2)：103-106,4.
基于AR和SVM的运动想象脑电信号识别
Recognition of motor imagery EEG based on AR and SVM
摘要
Abstract
A method of discrimination used for EEG signal of different movement imaginations was proposed. It propose auto-regressive parameter model coefficient and support vector machine could used to discriminate, 2008 BCI Competition Ⅳ Data set 2a were used in this paper. In this datasets, two types motor imagery EEG based on left hand and right hand imagery were extracted feature by auto-regressive parameter, and classified feature by support vector machine, different kernel functions were used to classify in compared experiment, the recognition accuracy can reach 75%.关键词
信号识别/脑-机接口/自回归/支持向量机/运动想象/分类识别Key words
signal recognition/ brain-computer interface (BCI)/ auto-regressive/ support vector machine (SVM) / motor imagery/ classification分类
信息技术与安全科学引用本文复制引用
张毅,杨柳,李敏,罗元..基于AR和SVM的运动想象脑电信号识别[J].华中科技大学学报（自然科学版）,2011,39(z2):103-106,4.基金项目
科技部国际合作资助项目(2010DFA12160) （2010DFA12160）
重庆市科技攻关项目(CSTC,2010AA2055). （CSTC,2010AA2055）
