LM算法在神经网络脑电信号分类中的研究

在脑机接口(BCI)中,脑电信号(EEG)的特征提取和分类识别可以通过多层前馈神经网络的大量学习来实现,但是基于误差反向传播的BP神经网络标准算法收敛速度慢,在训练中效率不高,分类正确率也很有限。针对这些问题,文中提出使用一种快速稳定的Levenberg-Marquardt算法来代替BP算法进行神经网络的学习训练,并利用BCI 2008竞赛的Graz数据集B进行了对左右手想象运动脑电信号分类的MATLAB仿真实验。该方法使得脑电信号分类的正确率达到87.1%,比BP算法的正确率78.2%要高,并且具有更好的收敛性。该算法为脑电信号的分类提供了有效的手段。     

一种新的BP学习算法及在运动想象脑电信号分类识别中的应用

对运动想象(MI)脑电信号的正确分类是决定基于运动想象脑电的脑－机接口(BCI)性能的关键因素。为有效地提取MI脑电信号特征、提高分类正确率,提出一种基于单形进化的BP神经网络优化算法(BPSSSE)并运用于MI脑电信号的识别,提取自相关(AR)模型参数和希尔伯特边际谱作为特征输入,通过单形进化算法优化BP神经网络学习性能,实现对MI脑电信号的分类。测试实验中,对BCI竞赛数据进行左右手分类。结果表明在4s~ 8s时间段内平均分类正确率为80.17%,最高分类正确率为87.14%,证明了本文算法在基于MI脑电的脑机交互控制系统中应用研究的有效性和可行性。     

基于小波和神经网络的脑电智能诊断研究

本文将数字信号处理、小波分析和神经网络技术结合起来应用到脑电智能诊断中.首先基于数字信号处理技术设计了两种滤波器:一是简单的FIR滤波器,二是格型滤波器.分别用来过滤工频伪迹和复杂的伪迹.然后利用小波信号处理技术,把已过滤的脑电信号分解到不同尺度中.最后基于在不同尺度下伪迹和异常波不会完全相同的原理,利用神经网络作为分类工具.把分解后的脑电信号输入神经网络进行识别,输出异常波的识别结果.     

基于EMD和PCA的P300分类算法

提出一种基于经验模态分解(EMD)及主分量分析(PCA)的分类算法,采用支持向量机(SVM)对P300脑电信号字符拼写实验进行分类,通过EMD变换对P300脑电信号分解,从而达到去噪增强特征的效果,使用PCA方法对原始P300信号进行特征提取和集中,并送入SVM中实现分类。实验结果表明,该算法能获得高达96%的分类正确率。     

基于EMD和PCA的P300分类算法

提出一种基于经验模态分解(EMD)及主分量分析(PCA)的分类算法,采用支持向量机(SVM)对P300脑电信号字符拼写实验进行分类,通过EMD变换对P300脑电信号分解,从而达到去噪增强特征的效果,使用PCA方法对原始P300信号进行特征提取和集中,并送入SVM中实现分类。实验结果表明,该算法能获得高达96%的分类正确率。     

基于经验模态－小波包分解和卷积神经网络的 Ｐ３００ 脑电信号分类研究

针对P300脑电信号信噪比低、随机性强及个体差异性大等问题,本文提出了一种将经验模态分解（EMD）和小波包分解（WP）相结合的滤波方法,并使用改进的卷积神经网络（CNN）对脑电信号进行分类识别。首先利用经验模态分解算法将原始脑电信号分解成若干个本征模函数（IMF）分量,并对每个分量进行频谱分析以去除主频段在0~30Hz以外的分量;然后,对保留的IMF分量进行小波包分解,根据P300电位的有效时频信息,选择合适的频段进行重构,再将重构后的各个本征模函数叠加,得到经过滤波后的脑电信号;最后,设计合适的卷积神经网络结构,对P300信号进行分类识别。本文使用国际BCI竞赛数据集对提出的方法进行验证。实验结果表明,两名被试的分类准确率分别为97.78%、95.56%,说明该方法能够有效的改善P300信号的识别效果（相比其他方法至少提升了2.78%,1.39%）,为进一步提高基于P300信号的脑机接口系统的性能提供了一种新的有效的途径。     

一种基于运动想象的脑机接口软硬件系统设计与实现

随着计算机科学、信号处理技术、模式识别算法等的飞速发展,脑机交互成为新的研究热点.本文设计一种少通道的脑机接口设备,并使用一种简便方法对脑电信号进行特征提取,保持了一定的分类准确率,有较强的工程意义.     

基于运动想象脑电信号的导联排序研究

为在脑机接口系统BCI(brain-computer interface)中有效选择导联进行特征提取和分类提供依据,研究了基于运动想象脑电信号的导联排序.根据公共空间模式算法CSP(common spatial pattern)原理提出了一种导联排序方法--基于协方差和主成分分析的排序算法CPSorting(covariance and principal component sorting),并研究了运动想象脑电信号MI(motor imagery)导联的排序情况以及排序靠前的导联对分类的贡献.利用公共空间模式算法对CPSorting排序后导联的数据提取特征,再分别应用支持向量机SVM和K近邻算法KNN进行分类.实验结果表明了该排序算法能有效地对基于运动想象脑电信号的导联进行排序.     

组合方法在脑电信号消噪中的仿真研究

研究脑电信号消噪问题.脑电信号存在非平稳性且包括大量的噪声,传统的消噪算法不能很好消除脑电信号中的噪声,从而影响后继的脑电信号处理和分析.为了更好的消除脑电信号噪声,提出一种小波变换与自适应滤波相结合的脑电信号组合消噪方法.该方法首先对含噪的脑电信号进行白化处理,然后采用小波分解和重构含噪较大的信号,将重构后的信号作为自适应滤波器的输入,进行自适应滤波消噪处理.仿真结果表明,组合去噪方法能有效去除脑电信号中的噪声干扰.     

改进孪生网络的脑电信号处理方法

针对运动想象脑机接口系统中分类准确率低的问题,提出一种改进孪生网络的脑电信号分类方法,把原孪生网络中的两个子网络扩充成三个子网络,并设计了新的学习样本采集方法和距离函数。脑电信号经过小波变换及经验模态分解,利用自相关系数筛选得到预处理后的小波分量,然后随机分割成训练集和测试集,从训练集中按照新的学习样本采集方法获得学习样本集,将其输入三个权重共享的子网络进行训练,使用新的距离函数进行相似度的对比,最后计算测试样本特征与训练集中标签为1和标签为0样本特征相似度,选择最高相似度样本标签作为该待测样本的类别。通过对国际公开BCI Competition II Data set III和The largest SCP data of Motor-Imagery数据集进行仿真,此算法分类准确率高达94.29%。与现有性能较高的算法进行对比,其有效的提高了分类准确率,能更好的进行脑电信号分类识别。     

Epileptic seizure classification in EEG signals using second-order difference plot of intrinsic mode functions

Epilepsy is a neurological disorder which is characterized by transient and unexpected electrical disturbance of the brain. The electroencephalogram (EEG) is a commonly used signal for detection of epileptic seizures. This paper presents a new method for classification of ictal and seizure-free EEG signals. The proposed method is based on the empirical mode decomposition (EMD) and the second-order difference plot (SODP). The EMD method decomposes an EEG signal into a set of symmetric and band-limited signals termed as intrinsic mode functions (IMFs). The SODP of IMFs provides elliptical structure. The 95% confidence ellipse area measured from the SODP of IMFs has been used as a feature in order to discriminate seizure-free EEG signals from the epileptic seizure EEG signals. The feature space obtained from the ellipse area parameters of two IMFs has been used for classification of ictal and seizure-free EEG signals using the artificial neural network (ANN) classifier. It has been shown that the feature space formed using ellipse area parameters of first and second IMFs has given good classification performance. Experimental results on EEG database available by the University of Bonn, Germany, are included to illustrate the effectiveness of the proposed method.     

基于样本熵和模式识别的脑电信号识别算法研究

脑-机接口BCI是一种实现人脑和外部设备通信的新兴技术。基于时频特性进行特征提取的传统方法无法体现EEG信号的非线性特征。为了进一步提高分类的准确率,首先采用小波阈值降噪的预处理方法提高了EEG信号的信噪比。然后结合非线性动力学的样本熵参数,对3种想象运动的脑电信号进行特征提取,保留了脑电信号的非线性特征。其中,运动想象MI脑电信号的研究一直都是BCI这一高速发展领域的重点目标。还研究了支持向量机、LVQ神经网络和BP神经网络3种分类器。通过实验结果对比发现,BP神经网络具有较高的识别率,更适用于脑电信号的分类识别。     

Detection of Seizure Activity in EEG by an Artificial Neural Network: A Preliminary Study

Neural networks, inspired by the organizational principles of the human brain, have recently been used in various fields of application such as pattern recognition, identification, classification, speech, vision, signal processing, and control systems. In this study, a two-layered neural network has been trained for the recognition of temporal patterns of the electroencephalogram (EEG). This network is called a Learning Vector Quantization (LVQ) neural network since it learns the characteristics of the signal presented to it as a vector. The first layer is a competitive layer which learns to classify the input vectors. The second, linear, layer transforms the output of the competitive layer to target classes defined by the user. We have tested and evaluated the LVQ network. The network successfully detects epileptiform discharges (EDs) when trained using EEG records scored by a neurologist. Epochs of EEG containing EDs from one subject have been used for training the network, and EEGs of other subjects have been used for testing the network. The results demonstrate that the LVQ detector can generalize the learning to previously “unseen” records of subjects. This study shows that the LVQ network offers a practical solution for ED detection which is easily adjusted to an individual neurologist's style and is as sensitive and specific as an expert visual analysis.     

结合概率型神经网络(PNN)和学习矢量量化(LVQ)算法的文本分类方法

针对文本自动分类问题,提出一种基于概率型神经网络(PNN)和学习矢量量化(LVQ)相结合的文本分类算法,该方法借助TFIDF方法提取文本特征及特征值,形成文本分类特征向量,利用概率型神经网络构建分类模型,并利用LVQ学习算法对神经网络模型竞争层网络进行学习,使相应模式向量相互靠拢,远离其他模式,从而实现文本分类.实验结果表明,提出的该方法在文本分类中表现了很好的效果,不仅具有很好的分类准确率,还表现出很好的学习效率.     

Discrete harmony search based expert model for epileptic seizure detection in electroencephalography

Seizure detection and classification using signal processing methods has been an important issue of research for the last two decades. In the present study, a novel scheme was presented to detect epileptic seizure activity with very fast and highest accuracy from background electro encephalogram (EEG) data recorded from epileptic and normal subjects. The proposed scheme is based on discrete wavelet packet transform (DWT) with energy, entropy, standard deviation, mean, kurtosis, skewness and entropy estimation at each node of the decomposition tree followed by application of probabilistic neural network (PNN). Normal as well as epileptic EEG epochs were decomposed into approximation and details coefficients till sixth-level using DWT packet. Discrete harmony search with modified differential operator was used to select the optimal features out of all above mentioned statistical and non-statistical parameters. In order to demonstrate the efficacy of the proposed algorithm for classification purpose using PNN, we have implemented 10-fold cross validation. Clinical EEG data recorded from normal as well as epileptic subjects are used to test the performance of this new scheme. It is found that the detection rate is 100% accurate with same level of sensitivity and specificity.     

基于经验模态分解和SVM的脑电信号分类方法

脑电信号的非线性、非平稳性造成对运动想象脑电信号的分类识别存在特征提取困难、可区分性低以及分类识别性能差等问题。本文提出一种基于经验模态分解(Empirical Mode Decomposition, EMD)和支撑向量机(Support Vector Machine, SVM)的运动想象脑电信号分类方法,充分利用EMD算法在处理非线性、非平稳信号的自适应性以及SVM在小样本条件的高识别性能和强泛化能力。首先利用EMD算法将C3、C4导联信号分解为一系列本征模函数(Intrinsic Mode Function, IMF),然后从IMF的信息和能量等维度提取特征将脑电信号转换至区分性更强的特征域,最后利用SVM进行分类识别。采用国际BCI竞赛2003中的Graz数据进行验证,所提方法可以得到94.6%的正确识别率,为在线脑-机接口系统的研究提供了新的思路。     

基于K-means聚类和LVQ神经网络的OLED缺陷像素识别

提出改进的K-means聚类分割和LVQ神经网络分类的方法，用于有机发光二极管显示面板喷墨打印制程中缺陷像素的识别。首先采用改进的K-means聚类算法对预处理后的打印像素进行分割，然后采用连通域水平矩形确定每一个打印像素的坐标及几何特征，再通过灰度共生矩阵提取其纹理特征，最后通过LVQ神经网络对所述特征进行分类，完成缺陷像素的标记及分类统计。结果表明，本文算法的识别率明显优于其他常用分类识别算法，平均缺陷检测率为100%，分类准确率达到98.9%，单像素检测时间为8.3 ms。     

An integrated SAR/GIS approach for investigating urban deformation phenomena: a case study of the city of Napoli,Italy

A classification method for polarimetric SAR data analysis using a competitive neural network is considered. The network is trained by two LVQ algorithms. In addition, a specific feature vector as the input for the network employing the JM distance is determined. As a result of experiments using SIR-C data, average accuracy for classification results was 86.40%, where (i) the competitive neural network with 8-input and 40-output neurons was trained by LVQ1 and LVQ2.1, and (ii) the 8-dimensional feature vector with backscattering coefficients (dB) and pseudo-relative phases between HH and VV from L and C bands was used. It is shown that the proposed method outperforms other methods in average accuracy.     

A self-adaptive data analysis for fault diagnosis of an automotive air-conditioner blower

This paper presents a fault diagnosis system for an automotive air-conditioner blower based on a noise emission signal using a self-adaptive data analysis technique. The proposed diagnosis system consists of feature extraction using the empirical mode decomposition (EMD) method and fault classification using the artificial neural network technique. The EMD method has been developed quite recently to adaptively decompose the non-stationary and non-linear signals. It sifts the complex signal of time series without losing its original properties and then obtains some useful intrinsic mode function (IMF) components. Calculating the energy of each component can reduce the computation dimensions and enhance classification performance. These energy features of various fault conditions are used as inputs to train the artificial neural network. In the fault classification, the probabilistic neural network (PNN) is used to verify the performance of the proposed system and compare with the traditional technique, back-propagation neural network (BPNN). The experimental results indicated the proposed technique performed well for quickly and accurately estimating fault conditions.     

基于小波变换和AdaBoost极限学习机的癫痫脑电信号分类

针对单一极限学习机(ELM)在癫痫脑电信号研究中分类结果不稳定、泛化能力差的缺陷,提出一种基于互信息(MI)的AdaBoost极限学习机分类算法。该算法将AdaBoost引入到极限学习机中,并嵌入互信息输入变量选择,以强学习器最终的性能作为评价指标,实现对输入变量以及网络模型的优化。利用小波变换(WT)提取脑电信号特征,并结合提出的分类算法对UCI脑电数据集以及波恩大学癫痫脑电数据进行分类。实验结果表明,所提方法相比传统方法以及其他同类型研究,在分类精度和稳定性上有着明显提高,并具有较好的泛化性能。