论文标题
最佳数据驱动雇用与代表性不足的组的权益
Optimal data-driven hiring with equity for underrepresented groups
论文作者
论文摘要
我们提出了一个以数据为基础的规范性框架，以进行公平决策，这是由雇用驱动的。雇主根据可观察的属性评估一组申请人。目的是聘请最好的候选人，同时避免对某些受保护的属性偏见。仅仅忽略受保护的属性不会由于数据中的相关性而消除偏差。我们提出了一项雇用政策，该政策取决于受保护的属性功能，但不是统计学上的属性，我们证明，在所有可能的公平政策中，我们的目标在公司的目标方面都是最佳的。我们在合成和真实数据上测试了我们的方法，并发现它显示出改善代表性不足和历史边缘化群体的股权的巨大实际潜力。
We present a data-driven prescriptive framework for fair decisions, motivated by hiring. An employer evaluates a set of applicants based on their observable attributes. The goal is to hire the best candidates while avoiding bias with regard to a certain protected attribute. Simply ignoring the protected attribute will not eliminate bias due to correlations in the data. We present a hiring policy that depends on the protected attribute functionally, but not statistically, and we prove that, among all possible fair policies, ours is optimal with respect to the firm's objective. We test our approach on both synthetic and real data, and find that it shows great practical potential to improve equity for underrepresented and historically marginalized groups.
