Fitnet: Can we forbid the normalization pre processing ?

Hello,

I would like to know if there is a possibility to avoid the pre processing included inside fitnet, rescaling the data between -1 and 1.

I want to feed it with my own scales/normalization method, is it possible ?

Answer 1

[Steven Lord]

net = fitnet(10);
net.inputs{1}

ans = 
    Neural Network Input

              name: 'Input'
    feedbackOutput: []
       processFcns: {'removeconstantrows', 'mapminmax'}
     processParams: {1x2 cell array of 2 params}
   processSettings: {1x2 cell array of 2 settings}
    processedRange: []
     processedSize: 0
             range: []
              size: 0
          userdata: (your custom info)

You want to modify the processFcns property of the net.inputs{1} object. See this documentation page for more information about the properties of that Neural Network Input object.