How to predict or forecast values using NFTOOL?

I am a new user MATLAB. I am working on an internship project titled "Short-Term Load Forecasting". After training the network on NFTOOL, I want to input some data and find the forecast values. Please have a look at the auto-generated script. Request Greg or some expert to help me find a solution.

% Solve an Input-Output Fitting problem with a Neural Network % Script generated by NFTOOL % Created Wed Jun 12 13:10:22 IST 2013 % % This script assumes these variables are defined: % % houseInputs - input data. % houseTargets - target data.

%filename = 'C:\Users\admin\Desktop\Matlab Study\Study\Testpurpose1.xlsx'; %sheet = 1; %xlRange1 = 'A2:C37'; %xlRange2 = 'D2:D37';

%inputs = xlsread(filename, sheet, xlRange1); inputs = xlsread('C:\Users\admin\Desktop\Matlab Study\Study\Testpurpose1.xlsx', 1, 'A2:C37'); targets = xlsread('C:\Users\admin\Desktop\Matlab Study\Study\Testpurpose1.xlsx', 2, 'A2:A37'); inputs = inputs'; %load house_dataset; %inputs = houseInputs; %targets = houseTargets; %targets = xlsread(filename, sheet, xlRange2); targets = targets';

% Create a Fitting Network hiddenLayerSize = 15; net = fitnet(hiddenLayerSize);

% Setup Division of Data for Training, Validation, Testing net.divideParam.trainRatio = 70/100; net.divideParam.valRatio = 15/100; net.divideParam.testRatio = 15/100;

% Train the Network [net,tr] = train(net,inputs,targets);

% Test the Network outputs = net(inputs); errors = gsubtract(targets,outputs); performance = perform(net,targets,outputs);

%p= [0.9,0.9,0.9]; I want to test such inputs and forecast the values. % View the Network view(net); %sim(net,p);

% Plots % Uncomment these lines to enable various plots. figure, plotperform(tr) figure, plottrainstate(tr) figure, plotfit(net,inputs,targets) figure, plotregression(targets,outputs) figure, ploterrhist(errors)

[Greg Heath]

1. Please format your post

2. The term forecast implies a time-series, a dynamic NN and the prediction of future outputs given past and/or current inputs. Do not use it if time delays are not involved. The preferred term for static NNs is predict.

3. I do not understand your problem... the house_dataset has a 13-dimensional input.