Adding new test dataset to Neural Network

I want to add a seperate test dataset into the Pattern recognition neural network. I have following datasets:

• input - 911*9 matrix with varius detailed information
• target - 911*2 matrix with two values either 1 or 0. 1 represents group A and 0 group B.
• test - 188*9 matrix with test data. We not know to which group it belongs.

Here is my code below but it doesn't work since the t or target values is 2*911 matrix and my y (where I'm adding my testing data) is 2*188 and the function e = gsubtract(t,y) always displayes the same error:

Error using bsxfun
Non-singleton dimensions of the two input arrays must match each other.

Error in gsubtract (line 22)
  c = bsxfun(@minus,a,b);

Can you please help me to add the test dataset into the neural network? Thank you!!

Here is my code (with errors...):

% This script assumes these variables are defined:
%   inputs - input data, size = [9,911]
%   target - target data, size = [2,911] filled with 1 and 0. 1 is group A and 0 is group B.
x = inputs';
t = target';

% Choose a Training Function
% For a list of all training functions type: help nntrain
% 'trainlm' is usually fastest.
% 'trainbr' takes longer but may be better for challenging problems.
% 'trainscg' uses less memory. Suitable in low memory situations.
trainFcn = 'trainlm';  % Scaled conjugate gradient backpropagation.

% Create a Pattern Recognition Network
hiddenLayerSize = 10;
net = patternnet(hiddenLayerSize);

% Choose Input and Output Pre/Post-Processing Functions
% For a list of all processing functions type: help nnprocess
net.input.processFcns = {'removeconstantrows','mapminmax'};
net.output.processFcns = {'removeconstantrows','mapminmax'};

% 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 = 30/100;
net.divideParam.testRatio = 0;

% Choose a Performance Function
% For a list of all performance functions type: help nnperformance
net.performFcn = 'crossentropy';  % Cross-Entropy

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

% Train the Network
[net,tr] = train(net,x,t);

% Test the Network
y = net(test');
e = gsubtract(t,y);
performance = perform(net,t,y)
tind = vec2ind(t);
yind = vec2ind(y);
percentErrors = sum(tind ~= yind)/numel(tind);

% Recalculate Training, Validation and Test Performance
trainTargets = t .* tr.trainMask{1};
valTargets = t .* tr.valMask{1};
testTargets = t .* tr.testMask{1};
trainPerformance = perform(net,trainTargets,y)
valPerformance = perform(net,valTargets,y)
testPerformance = perform(net,testTargets,y)

% View the Network
view(net)

Thank you!

Answer 1

[Greg Heath]

If you have test data inputs but do not know the corresponding correct outputs, the best you can do is estimate the output.

Without knowing the correct output you cannot calculate a figure of merit.

Hope this helps.

Thank you for formally accepting my answer

Greg

[Lily]

Thx Greg. When you say "calculate a figure of merit" what exactly do you mean?

Can I train and validate the neural network with one set of input data and than test the neural network with another seperate set of data (with of without the correcsponding target vector for that perticular testing dataset)?

[Greg Heath]

Yes. However, without the corresponding target, you cannot calculate a figure of merit like mse.