基于共空间模式与决策树支持向量机的运动想象脑电信号分类  

Classification of Motion Imagery EEG Signals Based on CSP and DTSVM

作　　者：张茜 耿晓中[2] 岳梦哲 汪林恩 户唯新 ZHANG Xi;GENG Xiaozhong;YUE Mengzhe;WANG Linen;HU Weixin(School of Information and Control Engineering,Jilin Institute of Chemical Technology,Jilin 132022,China;School of Computer Technology and Engineering,Changchun Institute of Technology,Changchun 130012,China)

机构地区：[1]吉林化工学院信息与控制工程学院,吉林吉林132022 [2]长春工程学院计算机技术与工程学院,吉林长春130012

出　　处：《软件工程》2023年第11期16-19,共4页Software Engineering

基　　金：吉林省科技厅项目(基于高精度脑电测量仪的脑电信号处理功能研发)。

摘　　要：针对脑电信号(Electroencephalogram,EEG)采集易受干扰导致EEG分类准确率低的问题,提出一种基于共空间模式(Common Spatial Pattern,CSP)与决策树支持向量机法(Decision Tree Support Vector Machine,DTSVM)相融合的运动想象脑电信号处理方法。首先利用CSP算法对运动想象的EEG特征值进行特征提取,其次运用线性判别分析法(Linear Discriminant Analysis,LDA)、自适应增强分类法(Adaptive Boosting,Adaboost)和决策树支持向量机法分别对特征进行分类,最后通过实验对比发现,利用决策树支持向量机进行分类的分类效果最佳,分类准确率最高可达到92.52%。This paper proposes a motion imagination EEG(Electroencephalogram)signal processing method based on the fusion of CSP(Common Spatial Pattern)and DTSVM(Decision Tree Support Vector Machine),to address the problem of low classification accuracy caused by interference in EEG signal acquisition.Firstly,the CSP algorithm is used to extract the features of the EEG feature values of motion imagination.Then,the features are classified by methods of LDA(Linear Discriminant Analysis),Adaboost(Adaptive Boosting)and DTSVM.Finally,through experimental comparison,it is found that using DTSVM for classification has the best classification effect,with a maximum classification accuracy of 92.52%.

关 键 词：脑电信号 共空间模式 支持向量机 决策树法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]