LabVIEW
Neural networks using Machine Learning Toolkit
@Nardimes wrote:
also i have identical problem ........ am doing a project on brain tumour classification as 0 and 1 which will represent cancerous and non cancerous mr image using nn ..... below attached the program but cant get the desired output
you did revive an old thread - usually, I would advise you to open a new thread.
but actually, I am quite happy, as I would have missed duetcat's very well example in message 4
no - your task is not identical to the task of the op.
moreover, your task is quite similar to the one presented in the ML Toolkits examples.
there's only one column in your class.csv - no need for a 2d array...
the number of your labels (20) does not match the number of rows in your input data (also 20)
don't do a trainingsset testset split on such a small dataset.
Rather do something like this to get familiar with the tools at hand:
your data in class 0 and class 1 is obviously very similiar - however the NN can separate class 0 from class 1 quite well with an accuracy of 75 %
have you counter-checked how well Tensorflow or Torch does perform on your dataset?
