摘要
已有许多基于EEG信号构建的压力检测模型,这些模型通过刺激材料诱导信号的产生,过程烦琐且难以保证标签的准确度。对此,提出了一种基于ECG信号生成标签用于训练EEG模型的方法。首先通过实验获取ECG数据进行线性拟合;然后同时采集ECG和EEG信号,用ECG生成压力数据并进行离散化;最后使用支持向量机和离散化后的数据训练模型,得到基于EEG信号的压力分类模型。在二分类任务下达到了90.16%的精度,表明了生成标签和EEG模型的有效性。
There are many stress detection models based on EEG signals.These models induce signal generation by stimulating materials.The process is cumbersome and it is difficult to ensure the accuracy of the label.In this regard,a method of generating tags based on ECG signals for training EEG models is proposed.Firstly,obtain ECG data through experiments for linear fitting;then,collect ECG and EEG signals at the same time,use ECG to generate stress data and discretization;finally use support vector machines and discretized data to train the model to obtain stress classification based on EEG signals model.The accuracy of 90.16%is achieved under the two-class classification task,which shows the effectiveness of the generated label and EEG model.
作者
林颖
LIN Ying(School of Information Engineering,Guangzhou Panyu Polytechnic,Guangzhou 511483,China)
出处
《现代信息科技》
2021年第9期93-95,99,共4页
Modern Information Technology
基金
广东省教育厅课题(2019GKTSCX069)。
