How to get chaos features from a set of signals to build a ML classifier based on it?

I have a set of only two categorise time series signals and my supervisor asked me to get the chaos features from it:

he told me first to get the hilbert huang transform, the get the chaos features from its output, but I coulden't fined how to do this

Answer 1

[Jaynik]

Hi Mahmoud,

The hht function can be used in MATLAB to obtain the Hilbert-Huang Transform. You will need to give the intrinsic mode function (IMF) as an input. For obtaining IMF, you can perform the Empirical mode decomposition on the time-domain signal using the emd function. Here is a sample code for the same:

% s is time-domain signal
imf = emd(s);

% Compute the Hilbert spectrum of the signal, fs is the sample rate
[hs,f,t] = hht(imf, fs);

You can refer the following documentation to read more about these functions:

• emd: https://www.mathworks.com/help/signal/ref/emd.html
• hht: https://www.mathworks.com/help/signal/ref/hht.html

Once you obtain the Hilbert spectrum, you can extract the chaos features. You may find the Chaotic Systems Toolbox available on File Exchange to be helpful: https://www.mathworks.com/matlabcentral/fileexchange/1597-chaotic-systems-toolbox

Hope this helps!