Hi Jacob,
The activation function of the output layer for the Neural Network Fitting app cannot be determined from the GUI alone, as far as my knowledge goes.
However, with a simple MATLAB script as attached below, you can determine the activation function and can change it.
The code is:
net = feedforwardnet;
disp(net.layers{end}.transferFcn);
Here, ‘purelin’ will be the output, and it means that the activation function is linear.
The neural net defined in the variable ‘net’ is the same neural net that you attached in the question, you can view it via code using the code line:
view(net);
Now, in order to change the activation function, you can simply assign the variable ‘net.layers{end}.transferFcn’ to your custom value. Please look into the documentation, https://www.mathworks.com/help/deeplearning/ref/feedforwardnet.html as setting the activation function might throw errors if not done correctly.
Hope this helps!
