基于EEG的睡眠数据的分类

Sleep classification based on EEG

睡眠研究对于人的身心健康和工作生活有着重要的意义。睡眠过程中不同状态的标注,是睡眠研究的一个基础工作。本文采用单通道的脑电信号数据,将数据输入到深度置信网络中进行特征表达和分类学习。通过利用39个晚上的睡眠数据进行测试,达到了82.26%的平均分类准确率。

The research of sleeping is important for our people＇s healthy, work and life. Labeling different stages of sleep data is the basic job in the study of sleep. In this paper, we used a single channel of EEG, and put the data into the deep brief networks to represent features and do the classification, the average accuracy is 82.86% based on 39 datasets.