MATLAB: Neural network – target dataset

neural networktarget data

I am doing project on sleep apnes (face recognition). I have to use artificial neural network for training and testing the dataset. For target dataset the values must be in 0s and 1s. how can i change my own target dataset into 0 and 1. Is there any code or something else from which i can get my correct target dataset to be trained. Pls help me with it.

I have attached my train and test data.

Thank you in advance

Best Answer

i suppose first you can normilize the train set and use it to normalize test set

meant=mean(train)

mint=min(train)

maxt=max(train)

data=(data-mint)/(mint-maxt)

this for features , same for outputs, i think you need regression network

https://ch.mathworks.com/help/matlab/ref/double.normalize.html

https://stackoverflow.com/questions/44147203/linear-regression-with-feature-normalization-matlab-code

https://www.researchgate.net/post/how_to_normalize_data_using_matlab_How_do_you_now_use_the_normalized_data_set_to_forecast_using_ann_on_matlab

Related Solutions

MATLAB: How to train a feedfordward neural network with error weights

 Use the command window
 help mse
 doc mse

Hope this helps.

Thank you for formally accepting my answer

Greg

MATLAB: How to use a personal function to normalization data in the neural network

See the documentation

>> help nnparam.normalization

and

>> doc nnparam.normalization

Hope this helps