Supply eval_sample_weight for fit in EarlyStoppingShapRFECV #144
Conversation
This file contains 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
- LGBM internally uses the sample weights for the training set eval_metric calls - while clf metrics such as rocauc or precision won't be impacted, during evaluation, log loss as eval_metric will result in never stopping if it is used without sample_weighting.
Matgrb
approved these changes
Apr 28, 2021
|
Same discussion as previously - personally I don't (want to) use it like that, but it may make sense to add it as an option in the future! |
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Outdated suggestions cannot be applied.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.
If using sample weighting for fitting in LGBM, one should also supply it for the evaluation set, otherwise the early stopping condition won't be reached when using
binary_log_lossas theeval_metric. The reason is that training sample weights may increase the training log loss to be generally larger than the validation loss, even though the validation loss stopped improving.Most other metrics were not affected, which is why this wasn't caught before.