matlab neural net NAR-time serious

I am using matlab neural net time serious to predict the further 48 days data,now i only get 8000 data about something.it's a matrix of 1*8000.i wanna know the data from 8001 to 8048. I had used NAR-time serious of ntstool.But I only get three net:net,netc and nets.how to predict? and i thought i got some useful information in the script generated in the gui as follow:

% Step-Ahead Prediction Network
% For some applications it helps to get the prediction a timestep early.
% The original network returns predicted y(t+1) at the same time it is given y(t+1).
% For some applications such as decision making, it would help to have predicted
% y(t+1) once y(t) is available, but before the actual y(t+1) occurs.
% The network can be made to return its output a timestep early by removing one delay
% so that its minimal tap delay is now 0 instead of 1.  The new network returns the
% same outputs as the original network, but outputs are shifted left one timestep.
nets = removedelay(net);
[xs,xis,ais,ts] = preparets(nets,{},{},T);
ys = nets(xs,xis,ais);
stepAheadPerformance = perform(net,ts,ys)

But i don't know how to deal with it.

[Greg Heath]

You are continually misspelling the word

 SERIES

Hope this helps.

Greg

Answer 1

[Greg Heath]

The removedelay command doesn't make much sense to me. Therefore I continually advise against using it.

Before designing a time-series prediction net I advise using the statistically significant delays (i.e., the delays of the statistically significant correlations) to determine the best nonegative input delays, ID, and positive feedback delays, FD, to use for design.

Search the NEWSGROUP and ANSWERS using

 greg nncorr

Hope this helps.

Greg

[DU HOUYI]

Really thanks for your response!

Oops,in fact I'm not a native English speaker.Please ignore my spelling and grammar mistakes.

As you say,It's important to determine the best nonegative input delays to get best prediction,but I was already using ntstool to test the delays to got the not so bad fitting pattern in gui.So my problem isn't to get best delays but to get a interface or algorithm or something else to lemme input the subsequent timesteps(i.e.,input x(8001:1:8048) to get corresponding y(8001:8048)).

I hope my shoddy English express meanings clearly.

<<

>>

[Greg Heath]

I have many posts on NARNET. Search both the NEWSGROUP and ANSWERS using the search words

 GREG NARNET

 GREG NARNET TUTORIAL

Hope this helps.

Greg

PS: It is best to consult these references in reverse chronological order