TL; DR These are some notes on calibration of surrogate loss functions in the context of machine learning. But mostly it is …