Facets of Disparate Impact: Evaluating Legally Consistent Bias in Machine Learning

Jarren Briscoe, Assefaw Gebremedhin
5/8/2025
cs.CYcs.LG

No official code found. Best implementation: jarrenbr/objective-fairness-index (1 star, MIT, last commit 2025-03-25).

Abstract

Leveraging current legal standards, we define bias through the lens of marginal benefits and objective testing with the novel metric "Objective Fairness Index". This index combines the contextual nuances of objective testing with metric stability, providing a legally consistent and reliable measure. Utilizing the Objective Fairness Index, we provide fresh insights into sensitive machine learning applications, such as COMPAS (recidivism prediction), highlighting the metric's practical and theoretical significance. The Objective Fairness Index allows one to differentiate between discriminatory tests and systemic disparities.

View on arXivView PDF

Code Implementations(1)

jarrenbr/objective-fairness-indexCommunity100%
MIT
10PythonAug 8, 20241 years agoMIT

Similar Papers

Compatibility of Fairness Metrics with EU Non-Discrimination Laws: Demographic Parity & Conditional Demographic Disparity

86% match2023

Analyzing Fairness of Classification Machine Learning Model with Structured Dataset

86% match2024

DispaRisk: Auditing Fairness Through Usable Information

86% match2024

Counterpart Fairness -- Addressing Systematic between-group Differences in Fairness Evaluation

85% match2023

A Catalog of Fairness-Aware Practices in Machine Learning Engineering

85% match2024

Fairness-Accuracy Trade-Offs: A Causal Perspective

85% match1 repo2024

Cite this paper

@article{briscoe2025facets,
  title  = {Facets of Disparate Impact: Evaluating Legally Consistent Bias in Machine Learning},
  author = {Jarren Briscoe and Assefaw Gebremedhin},
  year   = {2025},
  eprint = {2505.05471},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2505.05471v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2505.05471.svg)](https://codeofpaper.com/paper/2505.05471)

Discussion