I have trained my time series NARX NN using 5*7875 inputs and 1*7875 target values. How do I make predictions for 5*10 new input values ?
1 vue (au cours des 30 derniers jours)
Afficher commentaires plus anciens
I have trained my time series NARX NN using 5*7875 inputs and 1*7875 target values. How do I make predictions for 5*10 new input values ?
0 commentaires
Réponses (1)
the cyclist
le 29 Juil 2023
Modifié(e) : the cyclist
le 30 Juil 2023
You don't describe the code you used in any detail, so it is not possible to be certain, but if you did something like what is described in this documentation page to build your NN, then you can build a closed-loop network with
[netc,Xic,Aic] = closeloop(net,Xf,Af);
and make predictions with that network using
Yc = netc(XPredict,Xic,Aic)
where XPredict stores your new input values.
If that page doesn't help you, I suggest your share your code. Expecting us to guess from a one-sentence description is not the most efficient way to get answers here.
2 commentaires
the cyclist
le 30 Juil 2023
According to the documentation (specifically the part I quoted in my answer), it looks like
prediction_for_new_input_data = netc(new_input_data,Xic,Aic)
does what you want.
I have not used this toolbox, so I am just quoting the documenation.
Voir également
Catégories
En savoir plus sur Modeling and Prediction with NARX and Time-Delay Networks dans Help Center et File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!