Sample weights merged #2
Open
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
added 4 commits
May 24, 2018 11:25
1) normalize label fractions in the weights and allow skewing 2) introducing label-wise weighting 3) add weight and expected weight information in the history and in the plots 4) tab to spaces
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Outdated suggestions cannot be applied.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 includes the sample weights. I tested it before and after the inclusion with a skewed sample.
Before:
After:
There is a small effect on the MC training but should be OK. Could be also due to the fix introduced in the last commit that I believe increased the variance of the trainings (now the seed for splitting the samples is fixed).
Please have a look, but should be OK to merge