how to use NARX tool for a vectors of input to output?

I have a input of matrix sample N = 100 with time span of 30 seconds with discretization with dt = 0.01. Input of and for and the corresponding output for . How can I use the "NARX Neural Networks" for the above data. Please refer above DATA.mat file . How could the vector 100 samples are used for training and validate the trained model?

Answer 1

[Vatsal]

Hi,

To use a NARX (Nonlinear Autoregressive with Exogenous Inputs) neural network in MATLAB for modeling a system where you have vector inputs and outputs, you can follow these steps:

% Load your data

% Transform input and output into cell arrays of cell arrays
input = arrayfun(@(idx) x{idx}, 1:size(x, 1), 'UniformOutput', false);
output = arrayfun(@(idx) y3{idx}, 1:size(y3, 1), 'UniformOutput', false);

% Create the NARX network
inputDelays = 1:2;
feedbackDelays = 1:2;
hiddenLayerSize = 10;
net = narxnet(inputDelays,feedbackDelays,hiddenLayerSize);

% Prepare the data for training
[Xs,Xi,Ai,Ts] = preparets(net,input,{},output);

% Train the network
net = train(net,Xs,Ts,Xi,Ai);

% Test the network
outputs = net(Xs,Xi,Ai);
errors = gsubtract(outputs,Ts);
performance = perform(net,Ts,outputs);
view(net)

For more information on “NARX Neural Networks”, the following resources may be helpful:

https://www.mathworks.com/help/deeplearning/ref/narxnet.html

https://www.mathworks.com/help/deeplearning/ug/design-time-series-narx-feedback-neural-networks.html

I hope this helps!