基于神经网络集成的思维脑电信号分类方法

摘要

提出基于神经网络集成算法的思维脑电信号分类方法,采用BP神经网络为分类器,对用AR参数提取的思维脑电特征进行分类.为进一步提高BP神经网络的分类性能,采用Bagging算法对BP神经网络分类器进行加权投票,实验表明,提出的方法具有很好的分类效果.%Proposed EEG signal classification method based on BP neural network and Bagging algorithm, using BP neural network classifier, with the AR parameters of the extracted EEG features thought to classify. To further enhance the performance of BP neural network classification, using Bagging algorithm BP neural network classifier weighted voting, experiments show that the proposed method has a good classification rate.