Prediction using Neural Network with multiple external input variables

Hi I am trying to forecast rainfall using five external variables. I search the net for help and found this post (<http://www.mathworks.com/matlabcentral/answers/14970-neural-network-multi-step-ahead-prediction)helpful>, specially the answer given by Mark Hudson Beale. I have rainfall data from 1900 - 2014 and I want to predict rainfall for the next ten years (2015 - 2025). I have working codes up to test the network (given below). After that I need to predict rainfall for next ten years. I tried to get help from the codes given by Mark Hudson Beale, but I have error in the line Ai2 = mat2cell([zeros(10,2); Ai2],[10 1],ones(1,2));. I would greatly appreciate any help to solve this issue

load Rain;

X = [TMax,TMin,Evap,VP,Rad];

X = tonndata(X,false,false);

T = tonndata(Rain,false,false);

trainFcn = 'trainlm';

inputDelays = 1:2;

feedbackDelays = 1:2;

hiddenLayerSize = 10;

net = narxnet(inputDelays,feedbackDelays,hiddenLayerSize,'open',trainFcn);

[x,xi,ai,t] = preparets(net,X,{},T);

net.divideParam.trainRatio = 70/100;

net.divideParam.valRatio = 15/100;

net.divideParam.testRatio = 15/100;

% Train the Network

[net,tr] = train(net,x,t,xi,ai);

% Test the Network

y = net(x,xi,ai);

e = gsubtract(t,y);

performance = perform(net,t,y)

%Performance (last 120 predictions)

pred_test = (y(1391:end));

actual_test = (t(1391:end));

performance_test = perform(net,pred_test, actual_test)

RMSE = sqrt(performance_test)

Answer 1

[Greg Heath]

Search the NEWSGROUP and ANSWERS using subsets of

 greg nncorr narxnet multistep tutorial

because you have to find a suitable set of lags.

Hope this helps.

Thank you for formally accepting my answer

Greg