[ML] Scale regularisers for final train #1755
Merged
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 is labelled |
tveasey
added a commit
to tveasey/ml-cpp-1
that referenced
this pull request
Mar 25, 2021
As we move towards training for hyperparameter tuning on a small fraction of the data set and final training on more we will suffer issues with overfitting if we don't address the bias this introduces estimating regularisers. Interestingly, we already see a mismatch in train and test errors on larger data sets where we only use two-folds. I tested this correction, which is the one we use when we downsample, on a variety of data sets and we ended up with lower mismatch between train and test errors.
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Outdated suggestions cannot be applied.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.
As we move towards training for hyperparameter tuning on a small fraction of the data set and final training on more we will suffer issues with overfitting if we don't address the bias this introduces estimating regularisers. Interestingly, we already see a mismatch in train and test errors on larger data sets where we only use two-folds. I tested this correction, which is the one we use when we downsample, on a variety of data sets and we ended up with lower mismatch between train and test errors.