LabelSegmentedKFold cross-validation iterator #4709
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This PR implements a variant of KFold that ensures that the train and test split is disjoint w.r.t. a third-party label.
109bf38 to
cf47467
Compare
|
Is this different from #4444 ? |
|
@amueller: oh, I haven't seen it. thx. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR implements a variant of KFold that ensures that the train and test split is disjoint w.r.t. a third-party label. So far as I have seen all other Leave*Out methods either return all or a sample of all combinations of the third-party labels. However, if each party should be incorporated as test instance and the number of possible labels is huge, this k-fold variant is more efficient.
Example: Assume that we have 1.000.000 observations from 10.000 different users and we would like to estimate the generalization performance for new users. So, the train and test split should be disjoint, each user should occur once in the test set, and "leave one user out" would be intractable.
This is related to #4583.