Autoregression Multivariate Prediction with a NARX NN

Version 1.0.1 (434 KB) by H Sanchez
This code aims to perform a prediction of time sequence variable based on three external inputs.
301 Downloads
Updated 27 May 2021

View License

Aim
This code aims to perform a prediction of time sequence variable based on three external inputs. Having the external inputs at the time t predict the response at the time t. Having the inputs and response at time t what will be the response at the time t+1?
Output
net: An accurate open-loop shallow network able to predict the response given three external inputs. Use this network for validation or evaluate the response when you have known data; for example, you would like to test your network using a test set.
netc: A closed-loop network to predict the response variable at the time t given three external inputs at the time t; for example, you measure the input 1, 2 and 3 and would like to know what the response value will be at the time t.
nets:One step ahead prediction. Suppose now that at the time step t you have the three inputs and the response and you would like to know what would be the next response value at the time t+1 based on the measurement at the time t. For example, stock market; you have the volume, open, low and close value (response ) today and you would like to know the close (response) for tomorrow.

Cite As

H Sanchez (2026). Autoregression Multivariate Prediction with a NARX NN (https://nl.mathworks.com/matlabcentral/fileexchange/93000-autoregression-multivariate-prediction-with-a-narx-nn), MATLAB Central File Exchange. Retrieved .

MATLAB Release Compatibility
Created with R2021a
Compatible with any release
Platform Compatibility
Windows macOS Linux
Version Published Release Notes
1.0.1

Some small edits and change the file exchange icon.

1.0.0