damian_king asked . 2021-07-19

What work space values do i need to save separately to test Classification

What work space values do i need to save separately to test Classification of a number of voice emotion recognition neural networks and compare a new input against several to give a result?  

     TrainingHappyInput = NNHappyTrainingInput;
    TragetHappy= ones(1,1536);
 % set to Happy = 2
   TragetHappy=TragetHappy*2;
net = newff([min(TrainingHappyInput) max(TrainingHappyInput)],[14 1],{'tansig' 'purelin'},'traingd');
net.trainParam.epochs = 2900;   %Maximum number of epochs to train
    net.trainParam.goal = 0.01;    %Performance goal
    net.trainParam.lr = 0.01;       %Learning rate
    net.trainParam.min_grad=1e-10;  %Minimum performance gradient
    net.trainParam.show = 25;       %Epochs between displays
    net.trainParam.time = inf;      %Maximum time to train in seconds

    HappyTestset=TrainingHappyInput(400:700);

    NetOutputHappyTestData = sim(net,HappyTestset);
subplot(2,1,1), 
       plot(TrainingHappyInput(400:700),TragetHappy(400:700),HappyTestset,NetOutputHappyTestData,'o')
       title('Accuracy of classification'); 
  HappyDiffTraining =   TragetHappy (400:700)- NetOutputHappyTestData;
     subplot(2,1,2), plot(HappyDiffTraining);
     title('Difference Between Trained/Targets');

     HappyClassifiedTrained = mean(NetOutputHappyTestData);

     if HappyClassifiedTrained > 1.8078
         disp('Emotion detected is HAPPY...!');
     else 
         disp('Not Classified as Happy');
     end

neural network , classification , matlab , deep learning