Research map / Learning theory and algorithms / Fairness
Interpretable and Fair Boolean Rule Sets via Column Generation
Connor Lawless, Sanjeeb Dash, Oktay Gunluk, Dennis Wei
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 |
|---|---|---|---|
| use Hamming loss instead of 0-1 loss in the restricted MIP (ablation) | CG rule sets · adult | IP solve time (s): 278.5 -> 1.9 | Improved |
| use Hamming loss instead of 0-1 loss (ablation) | CG rule sets · bank-mkt | IP solve time (s): 165.6 -> 0.7 | Improved |
| use Hamming loss instead of 0-1 loss (ablation) | CG rule sets · FICO | test accuracy (%): 71.2 -> 71.0 | No clear effect |
| use Hamming loss instead of 0-1 loss (ablation) | CG rule sets · magic | test accuracy (%): 83.6 -> 83.4 | No clear effect |
| use CG instead of BRS | CG rule sets · tic-tac-toe | mean test accuracy (%): 99.9 -> 100.0 | No clear effect |
| use CG instead of AM | CG rule sets · tic-tac-toe | mean test accuracy (%): 84.3 -> 100.0 | Improved |
| use CG instead of RIPPER | CG rule sets · liver | mean test accuracy (%): 57.1 -> 59.7 | Improved |
| use CG instead of RIPPER | CG rule sets · gas | mean test accuracy (%): 99.0 -> 98.0 | Hurt |
| use CG instead of BRS | CG rule sets · gas | mean test accuracy (%): 92.2 -> 98.0 | Improved |
| use CG instead of RIPPER on complexity | CG rule sets · bank-mkt | complexity (# rules + conditions): 56.4 -> 9.9 | Improved |
Closest papers by meaning
- Boolean Decision Rules via Column Generation (NeurIPS 2018 · similarity 0.95)
- A Reductions Approach to Fair Classification (ICML 2018 · similarity 0.86)
- Learning Accurate and Interpretable Decision Rule Sets from Neural Networks (AAAI 2021 · similarity 0.85)
- Learning Certifiably Optimal Rule Lists for Categorical Data (JMLR 2018 · similarity 0.85)
- Efficient Fairness-Performance Pareto Front Computation (NeurIPS 2025 · similarity 0.84)
- A Sequentially Fair Mechanism for Multiple Sensitive Attributes (AAAI 2024 · similarity 0.83)
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.
