iqiukp/KPCA-MATLAB: MATLAB code for dimensionality reduction, feature extraction ...

原作者: [db:作者] 来自: 网络收藏邀请

开源软件名称（OpenSource Name）：

iqiukp/KPCA-MATLAB

开源软件地址(OpenSource Url)：

https://github.com/iqiukp/KPCA-MATLAB

开源编程语言(OpenSource Language)：

MATLAB 100.0%

开源软件介绍(OpenSource Introduction)：

Kernel Principal Component Analysis (KPCA)

MATLAB code for dimensionality reduction, fault detection, and fault diagnosis using KPCA

Version 2.2, 14-MAY-2021

Email: [email protected]

Main features

Easy-used API for training and testing KPCA model
Support for dimensionality reduction, data reconstruction, fault detection, and fault diagnosis
Multiple kinds of kernel functions (linear, gaussian, polynomial, sigmoid, laplacian)
Visualization of training and test results
Component number determination based on given explained level or given number

Notices

Only fault diagnosis of Gaussian kernel is supported.
This code is for reference only.

How to use

01. Kernel funcions

A class named Kernel is defined to compute kernel function matrix.

%{
        type   -
        
        linear      :  k(x,y) = x'*y
        polynomial  :  k(x,y) = (γ*x'*y+c)^d
        gaussian    :  k(x,y) = exp(-γ*||x-y||^2)
        sigmoid     :  k(x,y) = tanh(γ*x'*y+c)
        laplacian   :  k(x,y) = exp(-γ*||x-y||)
    
    
        degree -  d
        offset -  c
        gamma  -  γ
%}
kernel = Kernel('type', 'gaussian', 'gamma', value);
kernel = Kernel('type', 'polynomial', 'degree', value);
kernel = Kernel('type', 'linear');
kernel = Kernel('type', 'sigmoid', 'gamma', value);
kernel = Kernel('type', 'laplacian', 'gamma', value);

For example, compute the kernel matrix between X and Y

X = rand(5, 2);
Y = rand(3, 2);
kernel = Kernel('type', 'gaussian', 'gamma', 2);
kernelMatrix = kernel.computeMatrix(X, Y);
>> kernelMatrix

kernelMatrix =

    0.5684    0.5607    0.4007
    0.4651    0.8383    0.5091
    0.8392    0.7116    0.9834
    0.4731    0.8816    0.8052
    0.5034    0.9807    0.7274

02. Simple KPCA model for dimensionality reduction

clc
clear all
close all
addpath(genpath(pwd))

load('.\data\helix.mat', 'data')
kernel = Kernel('type', 'gaussian', 'gamma', 2);
parameter = struct('numComponents', 2, ...
                   'kernelFunc', kernel);
% build a KPCA object
kpca = KernelPCA(parameter);
% train KPCA model
kpca.train(data);

%　mapping data
mappingData = kpca.score;

% Visualization
kplot = KernelPCAVisualization();
% visulize the mapping data
kplot.score(kpca)

The training results (dimensionality reduction):

*** KPCA model training finished ***
running time            = 0.2798 seconds
kernel function         = gaussian 
number of samples       = 1000 
number of features      = 3 
number of components    = 2 
number of T2 alarm      = 135 
number of SPE alarm     = 0 
accuracy of T2          = 86.5000% 
accuracy of SPE         = 100.0000%

Another application using banana-shaped data:

03. Simple KPCA model for reconstruction

clc
clear all
close all
addpath(genpath(pwd))

load('.\data\circle.mat', 'data')
kernel = Kernel('type', 'gaussian', 'gamma', 0.2);
parameter = struct('numComponents', 2, ...
                   'kernelFunc', kernel);
% build a KPCA object
kpca = KernelPCA(parameter);
% train KPCA model
kpca.train(data);

%　reconstructed data
reconstructedData = kpca.newData;

% Visualization
kplot = KernelPCAVisualization();
kplot.reconstruction(kpca)

04. Component number determination

The Component number can be determined based on given explained level or given number.

Case 1

The number of components is determined by the given explained level. The given explained level should be 0 < explained level < 1. For example, when explained level is set to 0.75, the parameter should be set as:

parameter = struct('numComponents', 0.75, ...
                   'kernelFunc', kernel);

The code is

clc
clear all
close all
addpath(genpath(pwd))

load('.\data\TE.mat', 'trainData')
kernel = Kernel('type', 'gaussian', 'gamma', 1/128^2);

parameter = struct('numComponents', 0.75, ...
                   'kernelFunc', kernel);
% build a KPCA object
kpca = KernelPCA(parameter);
% train KPCA model
kpca.train(trainData);

% Visualization
kplot = KernelPCAVisualization();
kplot.cumContribution(kpca)

As shown in the image, when the number of components is 21, the cumulative contribution rate is 75.2656%，which exceeds the given explained level (0.75）.

Case 2

The number of components is determined by the given number. For example, when the given number is set to 24, the parameter should be set as:

parameter = struct('numComponents', 24, ...
                   'kernelFunc', kernel);

The code is

clc
clear all
close all
addpath(genpath(pwd))

load('.\data\TE.mat', 'trainData')
kernel = Kernel('type', 'gaussian', 'gamma', 1/128^2);

parameter = struct('numComponents', 24, ...
                   'kernelFunc', kernel);
% build a KPCA object
kpca = KernelPCA(parameter);
% train KPCA model
kpca.train(trainData);

% Visualization
kplot = KernelPCAVisualization();
kplot.cumContribution(kpca)

As shown in the image, when the number of components is 24, the cumulative contribution rate is 80.2539%.

05. Fault detection

Demonstration of fault detection using KPCA (TE process data)

clc
clear all
close all
addpath(genpath(pwd))

load('.\data\TE.mat', 'trainData', 'testData')
kernel = Kernel('type', 'gaussian', 'gamma', 1/128^2);
parameter = struct('numComponents', 0.65, ...
                   'kernelFunc', kernel);
               
% build a KPCA object
kpca = KernelPCA(parameter);
% train KPCA model
kpca.train(trainData);
% test KPCA model
results = kpca.test(testData);

% Visualization
kplot = KernelPCAVisualization();
kplot.cumContribution(kpca)
kplot.trainResults(kpca)
kplot.testResults(kpca, results)

The training results are

*** KPCA model training finished ***
running time            = 0.0986 seconds
kernel function         = gaussian 
number of samples       = 500 
number of features      = 52 
number of components    = 16 
number of T2 alarm      = 16 
number of SPE alarm     = 17 
accuracy of T2          = 96.8000% 
accuracy of SPE         = 96.6000%

The test results are

*** KPCA model test finished ***
running time            = 0.0312 seconds
number of test data     = 960 
number of T2 alarm      = 799 
number of SPE alarm     = 851

06. Fault diagnosis

Notice

If you want to calculate CPS of a certain time, you should set starting time equal to ending time. For example, 'diagnosis', [500, 500]
If you want to calculate the average CPS of a period of time, starting time and ending time should be set respectively. 'diagnosis', [300, 500]
The fault diagnosis module is only supported for gaussian kernel function and it may still take a long time when the number of the training data is large.

clc
clear all
close all
addpath(genpath(pwd))

load('.\data\TE.mat', 'trainData', 'testData')
kernel = Kernel('type', 'gaussian', 'gamma', 1/128^2);

parameter = struct('numComponents', 0.65, ...
                   'kernelFunc', kernel,...
                   'diagnosis', [300, 500]);
               
% build a KPCA object
kpca = KernelPCA(parameter);
% train KPCA model
kpca.train(trainData);
% test KPCA model
results = kpca.test(testData);

% Visualization
kplot = KernelPCAVisualization();
kplot.cumContribution(kpca)
kplot.trainResults(kpca)
kplot.testResults(kpca, results)
kplot.diagnosis(results)

Diagnosis results:

*** Fault diagnosis ***
Fault diagnosis start...
Fault diagnosis finished.
running time            = 18.2738 seconds
start point             = 300 
ending point            = 500 
fault variables (T2)    = 44   1   4 
fault variables (SPE)   = 1  44  18

鲜花

握手

雷人

路过

鸡蛋

该文章已有0人参与评论

请发表评论

全部评论

专题导读

More+

10-27 六六分期app的软件客服如何联系？(六六分期

11-06 可心卡盟:win10系统火狐flash插件崩溃怎么

11-06 亲亲特价:怎么删除回收站图标

11-06 济南大学虚拟社区:鲁大师节能降温的具体办

11-06 xlueops.exe:无线网络安装向导

11-06 女斗合众国:win7系统cf与主机连接不稳定怎

11-06 0xc000022-[cf烟雾头]cf怎么调烟雾头

11-06 qizideyouhuo:应用程序无法正常启动0xc0000

11-06 ipz-185:win7系统vcf文件怎么打开

11-06 傻哥蹦迪:win10系统s4怎么打开usb调试

11-06 八神浩树gtaste:回收站清空了怎么恢复

11-06 妖尾之黑色守护:win10系统电脑没有1440x900

11-06 校园至尊魔王小说:win7系统浏览网页时字体

11-06 女斗合众国:win10系统访问共享文件夹提示请

11-06 tokyo hot n0654:恢复win7系统默认字体一招

11-06 雨酷仙境:设置win7系统转移临时文件夹腾出

11-06 阿穆纳伊之杖:win7系统开始菜单在右边还原

11-06 tunespotting:win10系统火狐flash插件总是

11-06 甘尔葛分析师：计谋网站seo关键词暴涨有什

11-06 蔡贵霖: 计谋网站seo关键词暴涨有什么秘密

11-06 博益网首页:ao3网页版进入不了解决方法

11-06 漏斗子专栏: 网站数据分析小白易懂精华篇

11-06 见证双虹怎么做:win7系统开启telnet命令的

11-06 颾狐蝶蜋:系统资源不足无法完成请求的服务

11-06 国光中学校歌:提交网站到alexa查询详细步骤

11-06 西安有情天:静态网页和动态网页的区别

11-06 红木雅尚斋:外部链接构造对网站的好处

11-06 前官礼遇：防止域名劫持–增强域安全性的10

11-06 密传二转答案: 中文分词算法有哪些

11-06 金泉家园邮编:百度快照劫持的表现及应对方

cmontalvo251/MATLAB: This is my public repo of all my MATLAB Screen cast codes a ...发布时间：2022-08-17

danipascual/GNSS-matlab: Matlab codes to generate GNSS PRNs, secondary codes, da ...发布时间：2022-08-17

剪的笔顺,诠释剪的笔画,认识剪的部首

1 六六分期app的软件客服如何联系？(六六分期

六六分期app的软件客服如何联系？不知道吗？加qq群【895510560】即可！标题：六六分期

阅读：9412|2023-10-27

2 可心卡盟:win10系统火狐flash插件崩溃怎么

今天小编告诉大家如何处理win10系统火狐flash插件总是崩溃的问题，可能很多用户都不知

阅读：6628|2022-11-06

3 亲亲特价:怎么删除回收站图标

今天小编告诉大家如何对win10系统删除桌面回收站图标进行设置，可能很多用户都不知道

阅读：5521|2022-11-06

4 济南大学虚拟社区:鲁大师节能降温的具体办

今天小编告诉大家如何对win10系统电脑设置节能降温的设置方法，想必大家都遇到过需要

阅读：6010|2022-11-06

5 xlueops.exe:无线网络安装向导

我们在使用xp系统的过程中,经常需要对xp系统无线网络安装向导设置进行设置，可能很多

阅读：5877|2022-11-06

6 女斗合众国:win7系统cf与主机连接不稳定怎

今天小编告诉大家如何处理win7系统玩cf老是与主机连接不稳定的问题，可能很多用户都不

阅读：6251|2022-11-06

7 0xc000022-[cf烟雾头]cf怎么调烟雾头

电脑对日常生活的重要性小编就不多说了，可是一旦碰到win7系统设置cf烟雾头的问题，很

阅读：5832|2022-11-06

8 qizideyouhuo:应用程序无法正常启动0xc0000

我们在日常使用电脑的时候，有的小伙伴们可能在打开应用的时候会遇见提示应用程序无法

阅读：5301|2022-11-06

9 ipz-185:win7系统vcf文件怎么打开

今天小编告诉大家如何对win7系统打开vcf文件进行设置，可能很多用户都不知道怎么对win

阅读：5680|2022-11-06

10 傻哥蹦迪:win10系统s4怎么打开usb调试

今天小编告诉大家如何对win10系统s4开启USB调试模式进行设置，可能很多用户都不知道怎

阅读：5026|2022-11-06

客服电话

电子邮件

iqiukp/KPCA-MATLAB: MATLAB code for dimensionality reduction, feature extraction ...

开源软件名称（OpenSource Name）：

开源软件地址(OpenSource Url)：

开源编程语言(OpenSource Language)：

开源软件介绍(OpenSource Introduction)：

Kernel Principal Component Analysis (KPCA)

Main features

Notices

How to use

01. Kernel funcions

02. Simple KPCA model for dimensionality reduction

03. Simple KPCA model for reconstruction

04. Component number determination

05. Fault detection

06. Fault diagnosis

请发表评论

全部评论

上一篇：

下一篇：

CVE-2022-21531

dphi-official/Machine_Learning_Bootcamp

juven/maven-bash-completion: Maven Bash

win7系统注册表编辑器打开的操作方法

route101/mastoinker: Quick image view as

剪的笔顺,诠释剪的笔画,认识剪的部首

florent37/ViewAnimator: A fluent Android

florent37/Shrine-MaterialDesign2: implem

CVE-2020-36276

六六分期app的软件客服如何联系？(六六分期

doraiso/Mastodon

关于我们

产品与服务

解决方案

139-2527-9053