/Kernel-Principal-Component-Analysis-KPCA

KPCA for dimensionality reduction, feature extraction , fault detection, and fault diagnosis

Primary LanguageMATLAB

Kernel Principal Component Analysis (KPCA)

View Kernel Principal Component Analysis (KPCA) on File Exchange

MATLAB Code for non-linear dimensionality reduction, fault detection, and fault diagnosis through the use of kernels.

Version 2.1, 6-MAY-2020

Email: iqiukp@outlook.com

Main features

  • Easy-used API for training and testing KPCA model
  • Multiple kinds of kernel functions
  • Support for dimensionality reduction, fault detection, and fault diagnosis
  • Support for data reconstruction

Notices

  • Only fault diagnosis of Gaussian kernel is supported.
  • Class is defined using 'Classdef...End', so this code can only be applied to MATLAB after the R2008a release.
  • More details and discussions please see: https://www.ilovematlab.cn/thread-560380-1-1.html
  • This code is for reference only.

Demo for dimensionality reduction ('banana' data and 'circle' data)

Demo for data reconstruction ('circle' data)

Demo for fault detection (TE process data)

Demo for fault diagnosis (TE process data)