scikit-learn-contrib / hiclass

A python library for hierarchical classification compatible with scikit-learn
BSD 3-Clause "New" or "Revised" License
114 stars 20 forks source link

Add probabilistic classification to hiclass #minor #119

Closed LukasDrews97 closed 3 days ago

LukasDrews97 commented 7 months ago

Add probabilistic classification via calibration to hiclass using the following methods:

mirand863 commented 6 months ago

Hi @LukasDrews97,

Just a quick request from someone from France that reached out to me via e-mail. Would it be possible to add a threshold to remove labels that have low probability?

codecov-commenter commented 2 months ago

Codecov Report

Attention: Patch coverage is 90.49505% with 96 lines in your changes missing coverage. Please review.

Project coverage is 94.03%. Comparing base (4595264) to head (38c011a). Report is 4 commits behind head on main.

Files with missing lines Patch % Lines
hiclass/_calibration/VennAbersCalibrator.py 79.62% 43 Missing :warning:
hiclass/HierarchicalClassifier.py 85.84% 16 Missing :warning:
hiclass/metrics.py 95.45% 8 Missing :warning:
hiclass/LocalClassifierPerLevel.py 92.00% 6 Missing :warning:
hiclass/_calibration/Calibrator.py 91.89% 6 Missing :warning:
hiclass/Pipeline.py 44.44% 5 Missing :warning:
hiclass/LocalClassifierPerNode.py 95.78% 4 Missing :warning:
hiclass/_calibration/BetaCalibrator.py 87.87% 4 Missing :warning:
...iclass/probability_combiner/ProbabilityCombiner.py 88.88% 3 Missing :warning:
hiclass/LocalClassifierPerParentNode.py 98.76% 1 Missing :warning:
Additional details and impacted files ```diff @@ Coverage Diff @@ ## main #119 +/- ## ========================================== - Coverage 96.68% 94.03% -2.66% ========================================== Files 13 28 +15 Lines 1268 2297 +1029 ========================================== + Hits 1226 2160 +934 - Misses 42 137 +95 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.