Which of the following methods should a data scientist use just before switching to a potential replacement model?
A/B testing lets you compare the current model against the candidate in parallel, measuring performance on live data, before fully switching to the new model.
Delila
9 days agoAlex
10 days agoKara
10 days ago