MATLAB: How to organize the input and target samples for neural network

Deep Learning Toolboxneural network

Hi, i was preparing the input an target matrix of my dataset that i will used to create neural network using nprtool and i want to know does the input and target samples should be organized randomly? i menan if i have 2 Classes should i put the samples for the first class at the begining then i put the samples for the other class? because i load the example of cancer dataset and i find that the samples are put randomly

Best Answer

It doesn't make any difference. By default, the algorithm will randomize the order.

Hope this helps.

Greg

Related Solutions

MATLAB: How to test a trained network with new data

 trueclass        = vec2ind(target)  % Unit vector target columns --> class indices
 [net tr output ] = train(net,input,target);
 assignedclass    = vec2ind(output)
 err              = assignedclas~=trueclass
 Nerr             = sum(err)
 newoutput        = net(newinput)
 newassignedclass = vec2ind(newoutput)

Thank you for formally accepting my answer

Greg

MATLAB: Self Organizing Maps

The prototype pattern for each neuron is its weight vector. To see all the neurons' weight vectors:

net.IW

Each row represents the prototype vector for a different neuron.

These row/prototype vectors are what are shown in PLOTSOMPOS graphically, if the network has two or fewer inputs.

Best Answer

Related Solutions

MATLAB: How to test a trained network with new data

MATLAB: Self Organizing Maps

Related Question