Implementation of Nonlinear Principal Component Analysis for Time series data for time series data application

Good Day,

I would like be assisted with integrating a toolbox into Matlab software. The toolbox is called Nonlinear Principal Component Analysis(NLPCA). Here is a link to it:

http://www.nlpca.org/matlab.html. This toolbox implements a nonlinear version of linear Principal Component Analysis(PCA). PCA is radily available in Matlab, but i am interested in the Nonlinear version. This NLPCA is implemented by training a neural network in this toolbox. For my project, I want to implement this NLPCA on the timeseries data obtained from measuring instruments. I am not sure how to add this toolbox to My Matlab. I would appreciate the help I could get.

Secondly, I would like to be assisted with resources in relation to training a simple neural network on timeseries data. Even links would be appreciated. I learned that most people train neural networks on things like images. Not much on time series data.

Answer 1

[TC]

You can use this toolbox by simply downloading it and using it as a function in MATLAB.

Keep downloaded function in the directory (or path). and use [pc, net] = nlpca(Data, k) where Data is your dataset and k is the number of nonlinear components you want to extract. read its 'readme' file for more information.