How can I determine mean, standard deviation, minimum and maximum value of each type input and output in training, testing and validation data set, If i consider divide it randomly?

How can I determine mean, standard deviation, minimum and maximum value of each type input and output in training, testing and validation data set, If i consider Dividerand?

Answer 1

[Greg Heath]

[ net tr y e ] = train(net,x,t);

 Obtain trnind, valind and tstind from tr. To see how to do this, just
 type, without the ending semicolon,

 tr = tr

Hope this helps,

Greg

Thank you for formally accepting my answer

Greg

[grytan sarkar]

Thanks Greg. I have five inputs and one output. I want to see the mean, Std, min, max of each input in testing, validation and training set. It didnt works for me.

[Greg Heath]

How about showing what you tried?

[grytan sarkar]

Thanks Greg. I didnt use ind function

I have used mynet=feedforwardnet(hidden); mynet.trainFcn = 'trainlm'; mynet.performFcn='mse'; mynet.divideParam.trainRatio = 70/100; mynet.divideParam.valRatio = 15/100; mynet.divideParam.testRatio = 15/100;

Now I want to see which data is used in training, testing and validation? Also want to determine the mean, standard deviation of each input and output from randomly divided training, testing and validation data set. Although Mask{1} shows only data used in training, testing and validation set by 1 or NaN but I cant determine the mean or standard deviation from it.

[Greg Heath]

Reread my instructions:

The trn/val/tst indices are in the training record tr. Therefore, use the command

 tr = tr

and the rest will follow!

Greg

[grytan sarkar]

Thank you I got it. However how can I determine the standard deviation of inputs or outputs? std(tainInd) doest not work. Please suggest.