基于神经网络集成技术的运动想像脑电识别方法

An EEG Recognition Algorithm of Motor Imagery Based on Neural Network Ensemble

针对运动想像脑电信号,提出一种基于神经网络集成技术的识别方法.该方法通过离散小波变换(DWT)抽取想像左、右手运动的主要特征,采用Bagging算法构建神经网络集成(NNE)模型,并选取相对多数投票法获得网络集成的输出结果.在'BCI Competition 2003'竞赛数据集上的实验结果表明,该方法得到了比基于单个神经网络的脑电信号识别方法更高的识别率,同时,降低了个体神经网络的配置难度,提高了系统的泛化能力.

Being aiming at motor imagery EEG,a recognition algorithm based on Neural Network Ensemble （NNE） is presented.The main features of left-right hands imagery movement are obtained by using discrete wavelet transform（DWT）,then a NNE model is constructed by Bagging algorithm,and the output result of NNE is received through the relative majority vote method.The experiment results of BCI Competition 2003 database show that the recognition rate is better than that of the single neural network.Moreover,the configuration difficulty of single neural network is reduced,and the generalization ability of the system is improved.