quality_control: correct the per-class accuracy formula #7640
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
SpecLad
force-pushed
the
quality-formulas
branch
from
f09cb74 to
951d225
Compare
The current formula used to calculate `ConfusionMatrix.accuracy` is, in fact, not accuracy, but the Jaccard index. Replace it with the correct formula. Since the Jaccard index is a useful metric in its own right, calculate it too, but save it in another attribute of `ConfusionMatrix`.
SpecLad
force-pushed
the
quality-formulas
branch
from
951d225 to
60ba39e
Compare
Codecov Report
Additional details and impacted files@@ Coverage Diff @@
## develop #7640 +/- ##
========================================
Coverage 83.46% 83.46%
========================================
Files 373 373
Lines 39723 39725 +2
Branches 3741 3741
========================================
+ Hits 33154 33156 +2
Misses 6569 6569
|
zhiltsov-max
approved these changes
Mar 21, 2024
3 tasks
Merged
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Outdated suggestions cannot be applied.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.
Motivation and context
The current formula used to calculate
ConfusionMatrix.accuracyis, in fact, not accuracy, but the Jaccard index. Replace it with the correct formula.Since the Jaccard index is a useful metric in its own right, calculate it too, but save it in another attribute of
ConfusionMatrix.How has this been tested?
Manual testing.
Checklist
developbranch[ ] I have updated the documentation accordingly[ ] I have added tests to cover my changes[ ] I have linked related issues (see GitHub docs)[ ] I have increased versions of npm packages if it is necessary(cvat-canvas,
cvat-core,
cvat-data and
cvat-ui)
License
Feel free to contact the maintainers if that's a concern.