I have a binary classification problem. I have variables (features) var1, var2, var3, ..., var14.
Using these variables (aka features) in a logistic regression, I get their weights.
If I use the same set of variables in a neural network:
Should I get a different output?
Should I get the same output?
I developed a ROC Curve, and I have both lines overlaying on each other. I am not sure if I am missing something here.