Skip to content

ENH: Add Optimized Pre-Processing (Calmon et al.) (#9) - #34

Open
Chirag-agg wants to merge 1 commit into
usarfoss:mainfrom
Chirag-agg:feature/optimized-preprocessing
Open

ENH: Add Optimized Pre-Processing (Calmon et al.) (#9)#34
Chirag-agg wants to merge 1 commit into
usarfoss:mainfrom
Chirag-agg:feature/optimized-preprocessing

Conversation

@Chirag-agg

@Chirag-agg Chirag-agg commented Oct 15, 2025

Copy link
Copy Markdown

Description #9

This PR adds the Optimized Pre-Processing (OPP) algorithm from Calmon et al. to Fairlearn.

  • Performs demographic parity while limiting individual distortions.
  • Solves a linear program (LP) via cvxpy to compute optimized transformations for the features.
  • Provides a reference implementation for future preprocessing mitigation techniques.

Related issue: Closes #9.


Tests

  • no new tests required
  • new tests added (test/unit/preprocessing/test_optimized_preprocessing.py)
  • existing tests adjusted

Documentation

  • no documentation changes needed
  • user guide added or updated (docstring in OptimizedPreprocessing)
  • API docs added or updated
  • example notebook added or updated

Screenshots

N/A — this PR does not include visualizations.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

ENH Add mitigation algorithm from "Optimized Pre-Processing for Discrimination Prevention" by Calmon et al.

1 participant