[ML] Fix weights to maximize minimum recall for multiclass classification #1231
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
… objective on the sample set
tveasey
added a commit
to tveasey/ml-cpp-1
that referenced
this pull request
May 13, 2020
tveasey
added a commit
to tveasey/ml-cpp-1
that referenced
this pull request
May 13, 2020
tveasey
added a commit
that referenced
this pull request
May 13, 2020
tveasey
added a commit
that referenced
this pull request
May 13, 2020
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Outdated suggestions cannot be applied.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 is a correction for #1113.
We were computing the objective on the full training set rather than the sampled mask. Normally this is fine, since we use a stratified sample it basically acts like a constant scaling of the objective. However, when there is very little training data we can run into problems (such as a class missing in the sample set which is present in the training set). It also means this code is much slower than intended on very large data sets.
Change #1113 hasn't been released so I've marked this as a non-issue, but it must go out at the same time in 7.8.