New geometric framework analyzes robustness of AI fairness audits

Researchers have developed a new geometric framework to analyze the robustness of neighborhood-based fairness audits in machine learning. These audits assess individual fairness by comparing predictions for similar individuals in feature space. The new framework quantifies how small perturbations in feature space can alter local neighborhoods and lead to different fairness assessments, even if model predictions remain unchanged. It introduces a measure called 'audit volatility' to quantify the sensitivity of these audits under repeated perturbations, with experiments on benchmark datasets supporting the theoretical analysis. AI

IMPACT Provides a theoretical foundation for understanding and improving the reliability of fairness assessments in machine learning models.

RANK_REASON The cluster contains an academic paper detailing a new theoretical framework and experimental analysis for AI fairness audits. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• A Geometric Theory of Robust Fairness Audits
• arXiv
• audit volatility
• benchmark dataset
• computer science
• feature vector
• Hugging Face
• machine learning
• Neighborhood-based fairness audits

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →