摘要
为了提高脑机接口系统的分类准确率,论文对张量时频空模式分析特征提取算法进行了研究。该算法能够得到多维度上最佳分类特征,应用在模拟阅读脑-机接口脑电信号的特征提取中,与共空间模式相比,张量时频空模式能够得到更高的分类准确率。
In order to improve the classification accuracy of the brain-computer interface system, the feature extraction algorithm which called tensor temporal-frequency-spatial pattern is studied in this paper. This algorithm can extract the best classification feature in multi-dimensional. Comparing with common spatial pattern, tensor temporal-frequency-spatial pattern can get higher classification accuracy in Imitating-Reading brain-computer interface EEG signal feature extraction.
出处
《计算机与数字工程》
2015年第5期797-800,804,共5页
Computer & Digital Engineering
关键词
张量时频空模式
共空间模式
模拟阅读脑-机接口
tensor temporal-frequency-spatial pattern, common spatial pattern, imitating-reading brain-computer interface
