| 基于Adaboost技术的大脑运动意识任务分类 |
| |
| 引用本文： | 裴晓梅,郑崇勋. 基于Adaboost技术的大脑运动意识任务分类[J]. 北京生物医学工程, 2008, 27(5): 466-470 |
| |
| 作者姓名： | 裴晓梅 郑崇勋 |
| |
| 作者单位： | 西安交通大学生物医学工程研究所生物医学信息工程,教育部重点实验室,西安 710049 |
| |
| 基金项目： | 国家自然科学基金，中国博士后科学基金 |
| |
| 摘 要： | 本文提出了应用机器学习技术Adaboost算法与Fisher判别式分析作为基本分类器相结合的方法,实现大脑想象左右手运动意识任务的分类.利用Morlet小波滤波方法提取优化的运动相关脑电特征,对两组实验数据4个受试者运动相关脑电模式进行分类,平均最大分类正确率达到88.11%.通过最大分类正确率,最大互信息等评价指标比较,验证了Adaboost技术在改善大脑运动意识任务分类性能的有效性,从而为脑机接口系统应用中大脑运动意识任务分类提供了新的思路.
|
| 关 键 词： | 特征提取 Morlet小波滤波 互信息(MI) |
Classification of motor imagery tasks based on Adaboost technique |
| |
| PEI Xiaomei,ZHENG Chongxun. Classification of motor imagery tasks based on Adaboost technique[J]. Beijing Biomedical Engineering, 2008, 27(5): 466-470 |
| |
| Authors: | PEI Xiaomei ZHENG Chongxun |
| |
| Affiliation: | PEI Xiaomei,ZHENG Chongxun Key Laboratory of Biomedical Information Engineering of Education Ministry,Institute of Biomedical Engineering of Xi'an Jiaotong University,Xi'an 710049 |
| |
| Abstract: | The Adaboost classifier with fisher discriminant analysis(FDA) as base learner is proposed for discriminating the left and right hand motor imagery tasks in this paper.By Morlet wavelet filter to extract the optimized movement-related EEG features,the movement-related EEG patterns from two groups of experiment data including 4 subjects are distinguished and the average maximum classification accuracy reaches 88.11 percent.By the two indexes i.e maximum classification accuracy and mutual information(MI),the ... |
| |
| Keywords: | Adaboost ERD/ERS |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
