MATLAB CODE FOR ANN is not producing the desired output,program code is given .

inputs = [31 9650.00 3300.00 4350.00]; targets = [11 21 31 51];

% Create a Fitting Network hiddenLayerSize = 20; net = fitnet(hiddenLayerSize);

% Setup Division of Data for Training, Validation, Testing % For a list of all data division functions type: help nndivide net.divideFcn = 'dividerand'; % Divide data randomly net.divideMode = 'sample'; % Divide up every sample net.divideParam.trainRatio = 70/100; net.divideParam.valRatio = 15/100; net.divideParam.testRatio = 15/100;

% For help on training function 'trainlm' type: help trainlm % For a list of all training functions type: help nntrain net.trainFcn = 'trainlm'; % Levenberg-Marquardt

% Choose a Performance Function % For a list of all performance functions type: help nnperformance net.performFcn = 'mse'; % Mean squared error

% Choose Plot Functions % For a list of all plot functions type: help nnplot net.plotFcns = {'plotperform','plottrainstate','ploterrhist', ... 'plotregression', 'plotfit'};

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

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

% View the Network view(net) sim(net,outputs) % 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)

THE PROBLEM IS THAT I DONT GET OUTPUTS NEAR TO MY TARGETS

Answer 1

[Greg Heath]

The problem is that you do not have enough training data to sufficiently characterize a net of that size.

inputs = [31 9650.00 3300.00 4350.00]; 
targets = [11 21 31 51];
% Create a Fitting Network hiddenLayerSize = 20;

H=20

 [ I N ] = size(inputs)  % [ 1 4 ]
 [ O N ] = size(targets) % [ 1 4 ]
 Ntst = round(0.15^N)    % 1
 Nval = Ntst             % 1
 Ntrn = N-Nval-Ntst      % 2
 Ntrneq = N*O            % 2 = No. of training equations

CORRECTION:

 Ntrneq = Ntrn*O            % 2 = No. of training equations 

 Nw = (I+1)*H+(H+1)*O    % 61 = No. of unknown weights

Two equations are not enough to estimate 61 unknowns

[chaudhry]

sir thnx alot for the answer

sir kindly tell me Nw=(I+1)*h+(h+1)*o

from where this eq has come.. and this Nw eq stands for every code

and ntreq also

[Greg Heath]

Why didn't you search for the first few occurences in the NEWSGROUP and ANSWERS?

neural greg Nw

neural greg Ntrneq

How many training equations result from O output nodes and Ntrn input/target training pairs?

How many input weight/bias connections does each of h hidden nodes have from I input data nodes and 1 input bias node?

How many output weight/bias connections does each of O output nodes have from h hidden nodes and 1 output bias node?

Piece of cake! (just kidding)