Work

Training individually fair ML models with Sensitive Subspace Robustness

Year: 2019

Type: preprint

Abstract: We consider training machine learning models that are fair in the sense that their performance is invariant under certain sensitive perturbations to the inputs. For example, the performance of a resume screening system should be invariant under changes to the gender and/or ethnicity of the applicant. We formalize this notion of algorithmic fairness as a variant of individual fairness and develop a distributionally robust optimization approach to enforce it during training. We also demonstrate the effectiveness of the approach on two ML tasks that are susceptible to gender and racial biases.

Source: arXiv (Cornell University)

DOI: 10.48550/arxiv.1907.00020

Authors: Yurochkin, Mikhail, Bower, Amanda, Sun, Yuekai

Institutions: IBM (United States), University of Michigan

Language: English

Topic: Ethics and Social Impacts of AI

Subfield: Safety Research

Field: Social Sciences

Domain: Social Sciences

SDG: Reduced inequalities

Open Access status: green

Funder: National Science Foundation