How to predict next value using time series?
3 Ansichten (letzte 30 Tage)
Ältere Kommentare anzeigen
Hi, is there a way to use neural network toolbox to predict the next value in a time series?
0 Kommentare
Akzeptierte Antwort
Greg Heath
am 29 Okt. 2014
1. Override 'dividerand' to obtain constant intervals between points. For example
net.divideFcn = 'divideblock'
2. Obtain the final states Xf,Af from training
[ net tr Ys Es Xf Af] = train(net,Xs,Ts,Xi,Ai);
%Ys = net(Xs,Xi,Ai);
%Es = gsubtract(Ts,Ys);
3. You have to know the new initial state conditions Xinew, Ainew to predict points after the original data; Then
Ynew = net(NaN(n),Xinew,Yinew))
will yield n predictions. If Xinew,Ainew = Xf,Af, then Ynew is a gapless contnuation of Ys.
Hope this helps.
Thank you for formally accepting my answer
Greg
1 Kommentar
Weitere Antworten (2)
Greg Heath
am 24 Okt. 2014
Try
net = nar(1);
For details see
help nar
doc nar
Hope this helps
Thank you for formally accepting my answer
Greg
Star Strider
am 24 Okt. 2014
3 Kommentare
Star Strider
am 25 Okt. 2014
The targets are the predicted outputs for your training data. If you have no training targets, the net has nothing to adapt its weights with respect to. The correctly trained net will have the predictions as its outputs based on the input data.
Siehe auch
Kategorien
Mehr zu Sequence and Numeric Feature Data Workflows finden Sie in Help Center und File Exchange
Produkte
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!