A library hold-time model looks strong on the slice used to nudge hyperparameters. Which pile may be used for the statistical ML functional-performance test?
Select an answer to reveal the explanation.
Short Explanation
A library hold-time model that looks strong on the slice used to nudge hyperparameters still needs a large test dataset that is completely independent of training and validation. The nudge slice, the training pile, and any leftover carton that is not independent all leak the official score.
Full Explanation
ML functional performance testing requires an independent test dataset, not the training or validation pile. Using the nudge slice, the training pile, or any leftover carton misses that independence.