How to improve the accuracy of confusion matrix of neural network?

Hi, i used nprtool to create neural network for classification and i have dataset with input matrix 9*981 and target matrix 2*981. This is my code :

 x = inputpatient';
 t = targetpatient';
 % 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 = 'trainscg';  % Scaled conjugate gradient backpropagation.
 % Create a Pattern Recognition Network
hiddenLayerSize = 10;
net = patternnet(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,x,t);
 % Test the Network
y = net(x);
e = gsubtract(t,y);
performance = perform(net,t,y)
tind = vec2ind(t);
yind = vec2ind(y);
percentErrors = sum(tind ~= yind)/numel(tind);
 % View the Network
view(net)
 % Plots
% Uncomment these lines to enable various plots.
%figure, plotperform(tr)
%figure, plottrainstate(tr)
%figure, ploterrhist(e)
%figure, plotconfusion(t,y)
%figure, plotroc(t,y)

After training my net i got this confusion matrix

I want to improve my net because as you can see from my confusion matrix the accuracy is 65.9 % . Please i need help can anyone give me advice ?

Answer 1

[Greg Heath]

1. Search both the NEWSGROUP and ANSWERS using

 greg patternnet

and

 greg patternnet tutorial

2. Many of the posts

   a. Use as many defaults as possible
   b. Search for the smallest successful number of hidden nodes and 
      corresponding random initial weights using a double loop 
      approach:
        i) Outer loop over number of hidden nodes
       ii)  Inner loop over random initial weights.

Hope this helps.

Thank you for formally accepting my answer

Greg

[afef]

Thanks for your answer but i don't know how to do the part when you are talking about: "Search for the smallest successful number of hidden nodes and corresponding random initial weights using a double loop approach: i) Outer loop over number of hidden nodes ii) Inner loop over random initial weights." Can you please give me a hint?

[Greg Heath]

Hint: Searching with

         greg patternnet Ntrials

yields the following number of hits

NEWSGROUP    ANSWERS    GOOGLE    
   11          74       38,600

Hope this helps

Greg