Fix problem with NarX network

2 visualizaciones (últimos 30 días)
FRANCISCO
FRANCISCO el 20 de Mzo. de 2013
Very good, could rise as a snapshot or to explain the problem presented to me?. I made a prediction closeloop NarX, ie inputs have introduced today and given me the values of tomorrow. The problem is that when I look at the diagram in the original outputs (I do it for comparing the outputs with the outputs predicted original) gives me the answer 2 steps later. That is, if the original data make me a peak in January, I predicted outputs do in 3. I've thought about making remove delay, but I have a problem with the command since I required preparets arrays of inputs and targets of the same size after applying tonndata and I have the same dimensions as if tomorrow I get the targets, which now I have no inputs are the morning and therefore would not identical arrays. How I can solve the problem? thank you very much

Respuestas (0)

Categorías

Más información sobre Deep Learning Toolbox en Help Center y File Exchange.

Etiquetas

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by