Considering PIO + the UR template, what are some of the ways to score the model? I can think of some A/B testing but that would imply relying on things outside the PIO + UR stack. Is there a way to measure lift or "score the model"? Cross-validation with separate testing set?
Thanks Gustavo
