How to add HOG feature to the following code

Here is the code for the neural network, "Latest.mat" contains training and test images for the neural network,

• I need to see the accuracy and
• I need to add "HOG feature" and using that i need to compare between the training and the test images, Please can anyone help me with that please

clc; clear all;

%%

load latest.mat; trainFcn = 'trainscg'; % Scaled conjugate gradient backpropagation.

net.performFcn = 'mse'; %newly added

hiddenLayerSize = 100; %net = patternnet(hiddenLayerSize); net=feedforwardnet([20 10]);

view (net); net.divideParam.trainRatio = 70/100; net.divideParam.valRatio = 15/100; net.divideParam.testRatio = 15/100;

net.trainParam.epochs = 100;

[net,tr] = train(net,training,test); y = net(training); e = gsubtract(test,y); performance = perform(net,test,y) tind = vec2ind(test); yind = vec2ind(y); percentErrors = sum(tind ~= yind)/numel(tind);

%net.IW{1,1}; %net.b{1}; %net.LW {1};

% View the Network view(net)

Thank You

[Greg Heath]

1. Please use the {}Code operator to properly format your code for ease of reading.

2. I and thousands of other neural net users have no idea what the acronym HOG stands for.

3. Why don't you have validation subset images to help prevent overtraining?

Hope this helps

Greg

[Tousif Ahmed]

HOG stands for Histogram of oriented gradients, in the latest.mat file i have training and test data. I need to add HOG for comparison of data and to check accuracy

[Greg Heath]

3. Why don't you have validation subset images to help prevent overtraining?

[Tousif Ahmed]

Sorry i dont know much about that, can you please help me, i am new to matlab.

[Greg Heath]

The fudamental role of the validation subset is clearly explained in the documentation.

You can also search both the NEWSGROUP and ANSWERS using the search word

 validation

Hope this helps.

Greg

[Tousif Ahmed]

https://in.mathworks.com/matlabcentral/answers/58761-nn-validation-and-data-partition

I followed the above link..

Can you please help me in my code with the validation set please