NARX network-how to build a model

Hi,

I have a large data that I have to predict and simulate, e.g 1x30000. My idea is to take every 10th point or 5th of this large data and then complying with the large data. For example; y(t-10) can be used instead of y(t-1) in narx network, I meant the first point in the small data will be exactly the same at 10th point in the large data. After this idea I have implemented a narx network with LM algorithm (the default algorithm) and got a good result for this small data, but I cannot comply with the large data. Does anyone know how to do it or if it is possible or not? If it is possible can you tell me how I can do it? Regards

[mucagd]

no idea?