Algorithmic fairness with monotone likelihood ratios
Authors: Wes Camp
TMLR 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We show that inequalities of many commonly used fairness metrics (true/false positive/negative rates, predicted positive/negative rates, and positive/negative predictive values) are guaranteed for groups with different outcome rates under a monotonically calibrated model whose risk distributions have a monotone likelihood ratio, extending existing impossibility results. We further provide lower bounds on the FNR/FPR disparities and PPR/PNR disparities in the same setting, showing that either the FNR disparity or FPR disparity is at least as large as the positive outcome rate disparity (for FNR disparity) or negative outcome rate disparity (for FPR disparity), and either the PPR disparity or PNR disparity is at least as large as the positive outcome rate disparity (for PPR disparity) or negative outcome rate disparity (for PNR disparity). |
| Researcher Affiliation | Industry | Wes Camp EMAIL Optum |
| Pseudocode | No | The paper presents theoretical results, definitions, lemmas, and theorems. It does not include any explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | No | The paper focuses on theoretical analysis and does not describe a new method or system for which code would typically be released. There is no statement about making code available. |
| Open Datasets | Yes | Using Pro Publica s COMPAS data (Larson et al., 2016) Larson, Jeff, Surya Mattu, Lauren Kirchner, and Julia Angwin. How We Analyzed the COMPAS Recidivism Algorithm. 2016. URL https://www.propublica.org/article/ how-we-analyzed-the-compas-recidivism-algorithm. |
| Dataset Splits | No | The paper analyzes existing data from the COMPAS algorithm to illustrate theoretical findings. It does not conduct new experiments requiring specific training/test/validation splits for a model developed by the authors. |
| Hardware Specification | No | The paper is theoretical in nature and analyzes an existing algorithm (COMPAS) using publicly available data. It does not describe any new experimental setup, model training, or inference that would require specific hardware specifications. |
| Software Dependencies | No | The paper is theoretical, presenting mathematical proofs and analysis. It does not describe any specific software or libraries with version numbers used for implementing or running new experiments. |
| Experiment Setup | No | The paper is theoretical and provides proofs and analysis of fairness metrics. While it uses the COMPAS dataset as an example, it does not describe an experimental setup, hyperparameters, or training configurations for a model developed or trained by the authors. |