Classification loss and neural neural network

Hello,

I want to compare my neural network and my support vector machine with one another. For fitcsvm I am using the loss function loss(Mdl,XTest,YTest,'LossFun','exponential'). Now I want to calculate the exponential loss for my neural network. Is there a function which calculates the exponential loss for neural networks? Since I couldn´t find one I wanted to write my own, but I have difficulties to understand the math behind the implemented loss function (https://de.mathworks.com/help/stats/compactclassificationensemble.loss.html#bswgv2a-3).

How are the m and the w calculated?