Kappa系数在两类运动想象EEG识别中的应用考察
首发时间:2018-07-25
摘要:本文考察了kappa系数在衡量两类运动想象脑电信号识别效果方面的适用性及相关特点。本文以国际脑机接口竞赛的多个数据集作为研究样本,使用多种识别方法,包括Naive Bayes方法以及SVM方法,对脑电信号进行识别。通过对比得出,kappa系数适用于衡量两类运动想象脑电信号的识别效果,该参数能够充分反映出识别效果的变化。本文也指出,评价生理信号的识别效果是一个较为复杂的问题,而kappa系数则可以简化该问题,是较为实用的参数。
关键词: 生物电气接口 脑电 kappa系数 贝叶斯 支持向量机
For information in English, please click here
Investigation on the Application of Kappa Coefficient in Two-type Motor Imagery EEG Identification
Abstract:The paper investigated the applicability and relevant characteristic of kappa coefficient in two-type motor imagery EEG identification. In the paper, several EEG data sets of the international BCI competitions were used, and several identification methods including Naive Bayes method and SVM method were used to identify EEG. Through comparison, it was concluded that the kappa coefficient was suitable for measuring the identification effect of two-type motor imagery EEG, and this parameter can fully reflect the variation of identification effect. The paper also pointed that it was a complicated problem to evaluate the identification effect of physiological signals, while kappa coefficient can simplify this problem, hence, this parameter should be a practical parameter.
Keywords: BEI (Biotic Electric Interface) EEG (electroencephalogram) kappa coefficient Bayes SVM (support vector machine)
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
Kappa系数在两类运动想象EEG识别中的应用考察
评论
全部评论0/1000