# Fieldtrip:ICA remove the EOG

English version: Thttp://www.fieldtriptoolbox.org/example/use_independent_component_analysis_ica_to_remove_ecg_artifacts 下面用中文对这个教程走一遍 首先用ICA去除EOG需要四部曲： preparing MEG data for running an ICA  准备数据 decomposition of the MEG data                 解分数据 identifying the components that reflect heart artifacts 识别心电 removing those components and backprojecting

# Fieldtrip ft_definetrial and ft_redefinetrial?

Today when i came acrocss the ft Tutorial , there are two functions maybe confused me, ft_definetrial defines the segments of data that will be used for further processing and analysis, i.e. the pieces of data that will be read

# BP算法最接地气2

Principles of training multi-layer neural network using backpropagation The project describes teaching process of multi-layer neural network employingbackpropagation algorithm. To illustrate this process the three layer neural network with two inputs and one output,which is shown in the picture below,

# 最接地气的BP算法说明1Calculus on Computational Graphs: Backpropagation

Introduction Backpropagation is the key algorithm that makes training deep models computationally tractable. For modern neural networks, it can make training with gradient descent as much as ten million times faster, relative to a naive implementation. That’s the difference between

# EEG reference ? you should know or may know ?

EEG reference is the basic of operation for signal analysis , since you would like to reduce the noise greatfully, the problem is that you have seen so many kinds of technology to do the rerefence such like average common(all

# FieldTrip and EEGlab

Spend a lot of time on EEGlab and try to use the fieldtrip for the new Development of EEG analysis. EEGlab is a very nice gui toolbox in Matlab, the problem is that it is not so comprehensive for later

# statistic theory: Canonical Correlation analysis CCA and Linear discrimindate analysis LDA

X={x1+x2+….xn};   xm平均数 Y={y1+y2+…..yn};  ym平均数 方差 variance=（x1-xm）^2+(x2-xm)^2+……(xn-xm)^2  表示数据的离散程度 标准差 standern variance=variance^1/2                          方差的开方根标准差或均方差 协方差 衡量两个变量之间的关系： 在概率论中，两个随机变量 X 与 Y 之间相互关系，大致有下列3种情况： 当 X, Y 的联合分布像上图那样时，我们可以看出，大致上有： X 越大  Y 也越大， X

# 小波分析：信号的时频分析和幅频分析

Wavelet 是蛮复杂的一个信号分析模式，我们主要利用这个做EEG信号的时频分析 时频分析上述讲过可以通过短时间傅里叶变换来做，但是会造谷频率泄露 所以小波分析很好地解决这个问题，即可获得信号频率而且可以知道频率所在的位置 这篇主要分两个部分， 第一个是利用小波分析来画时频图 第二个是利用希尔伯特变换求瞬时频率 相位 还有振幅 @小波变换画时频图 小波有很多种，也可以参考窗函数，其实就是各种小波基 目前效果最高的应属morlet小波，其中心频率越大，体现出的效果越好   三个matlab函数 COEFS = cwt(S,SCALES,’wname’)     说明：该函数能实现连续小波变换，其中S为输入信号，SCALES为尺度，wname为小波名称。     FREQ = centfrq(‘wname’)     说明：该函数能求出以wname命名的母小波的中心频率。     F = scal2frq(A,’wname’,DELTA)     说明：该函数能将尺度转换为实际频率，其中A为尺度，wname为小波名称，DELTA为采样周期。      设a为尺度，fs为采样频率，Fc为小波中心频率，则a对应的实际频率Fa为