Research map / Learning theory and algorithms / Prediction and decision losses
Fair Wrapping for Black-box Predictions
Alexander Soen, Ibrahim Alabdulmohsin, Oluwasanmi O Koyejo, Yishay Mansour, Nyalleng Moorosi, Richard Nock, Ke Sun, Lexing Xie
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Wrap the black-box posterior with an input-dependent α-tree | black-box classifier · Toy hiring task | worst-case loss: 1.09 to 0.62 | Improved |
| Wrap the black-box posterior with an input-dependent α-tree | black-box classifier · Toy hiring task | CVaR β=0.9: 1.09 to 0.62 | Improved |
| Apply conservative or audacious α-tree updates | clipped calibrated RF · ACS 2015 CA income, binary age attribute | CVaR | Improved |
| Use audacious rather than conservative updates (ablation) | α-tree wrapper · ACS 2015 CA income, binary age attribute | CVaR | Improved |
| Use audacious rather than conservative updates (ablation) | α-tree wrapper · ACS 2015 CA income, binary age attribute | equality of opportunity | Hurt |
| Use audacious rather than conservative updates (ablation) | α-tree wrapper · ACS 2015 CA income, binary age attribute | meandiff | Hurt |
| Use conservative α-tree updates rather than the equality-of-opportunity baseline | α-tree wrapper · ACS 2015 CA income, binary age attribute | equality of opportunity | Improved |
| Use the equality-of-opportunity baseline rather than conservative α-tree updates | derived predictor baseline · ACS 2015 CA income, binary age attribute | meandiff | Improved |
| Apply conservative α-tree updates for statistical parity | clipped calibrated RF · ACS 2015 CA income, binary age attribute | statistical parity | Mixed |
| Use audacious rather than conservative updates (ablation) | α-tree wrapper · ACS 2015 CA income, binary age attribute | meandiff | Improved |
Closest papers by meaning
- Optimal Black-Box Reductions Between Optimization Objectives (NeurIPS 2016 · similarity 0.86)
- Fair Densities via Boosting the Sufficient Statistics of Exponential Families (ICML 2023 · similarity 0.85)
- Fair and Optimal Prediction via Post-Processing (AAAI 2024 · similarity 0.85)
- Fair Multiple Decision Making Through Soft Interventions (NeurIPS 2020 · similarity 0.85)
- Understanding Fairness and Prediction Error through Subspace Decomposition and Influence Analysis (NeurIPS 2025 · similarity 0.85)
- Reconciling Predictive and Statistical Parity: A Causal Approach (AAAI 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
