Validation is not inherently tied to repetition over folds. Considering the hundreds of "unanswered" questions tagged with [cross-validation] (meaning no answer is either up-voted or accepted), I propose the following new design:

Name: Statistics

Symbol: A Cauchy pdf overlaid with Gaussian, showing fat tails of former.


1 Answer 1


I take the name to mean that we do not just check one source (many members can answer), much as cross validation, as opposed to just checking on one out-of-sample set, checks many out-of-sample sets.

I would have voted for "Confidence Region" eleven years ago, but I like the name we have and see no reason to change.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .