算法公平性的秩毕业度量
A Rank Graduation metric for Algorithmic fairness
浏览论文内容
中文总结 AI 辅助
本文提出基于秩的公平性评估框架，结合RGF、AURGF、置换检验和特征移除，在模拟和HMDA数据上验证，发现树模型兼顾准确性与公平性，且不公平源于数据固有群体差异。
中文摘要 AI 辅助
在影响个人（如信用评分）的算法决策中，公平性评估通常依赖于在总体群体层面计算的奇偶性度量。此类度量可能无法揭示哪些个体遭受不公平，或哪些解释性因素导致了不公平。在本文中，我们提出一个基于秩的框架，通过模型预测误差的分布来评估公平性，从而将公平性评估与预测准确性和可解释性联系起来。该框架结合了秩毕业公平性（RGF）、其集成度量AURGF、中心化的Cramer--von Mises置换检验，以及用于公平性可解释性的特征移除程序。我们使用逻辑回归、随机森林、梯度提升和多层感知器评估该方法。模拟研究表明，受保护群体的不平衡可以逆转描述性公平性比较，而所提出的推断程序能正确区分公平与不公平机制。将其应用于HMDA抵押贷款数据产生的模型排名与使用经典公平性标准获得的排名不同。基于树的模型，而非逻辑回归，提供了预测准确性和基于秩的公平性的最强组合，而公平性零假设对所有四个模型均被拒绝。跨统计、装袋、提升和神经网络规格的差异持续性，连同特征移除结果，表明观察到的不公平并非特定于单一算法或预测器，而是与贷款数据特征中嵌入的群体差异相关。这些发现支持一种更广泛的值得信赖的人工智能方法，该方法结合了预测准确性、公平性度量、统计推断和可解释性。
英文摘要
Fairness assessment in algorithmic decisions that affect individuals, such as credit scoring, often relies on parity measures calculated at the aggregate group level. Such measures may not reveal which individuals experience unfairness or which explanatory factors contribute to it. In this paper, we propose a rank-based framework that evaluates fairness through the distribution of model prediction errors, thereby linking fairness assessment with predictive accuracy and explainability. The framework combines Rank Graduation Fairness (RGF), its integrated measure AURGF, a centered Cramer--von Mises permutation test, and a feature removal procedure for fairness explainability. We evaluate the methodology using logistic regression, random forest, gradient boosting, and a multilayer perceptron. The simulation study shows that protected-group imbalance can reverse descriptive fairness comparisons, whereas the proposed inferential procedure correctly distinguishes fair from unfair mechanisms. Its application to HMDA mortgage data produces model rankings that differ from those obtained with classical fairness criteria. Tree-based models, rather than logistic regression, provide the strongest combination of predictive accuracy and rank-based fairness, while the fairness null hypothesis is rejected for all four models. The persistence of disparity across statistical, bagging, boosting, and neural network specifications, together with the feature removal results, indicates that the observed unfairness is not specific to a single algorithm or predictor, but is associated with group differences embedded in the characteristics of the lending data. These findings support a broader approach to trustworthy artificial intelligence that combines predictive accuracy, fairness measurement, statistical inference, and explainability.
发表机构
- University of Tipaza（蒂帕扎大学）
- University of Pavia（帕维亚大学）
机构由 AI 辅助整理，请以论文原文为准。
