pith. sign in

arxiv: 2011.03173 · v2 · pith:ZOY6U46Snew · submitted 2020-11-06 · 📊 stat.ML · cs.LG

Does enforcing fairness mitigate biases caused by subpopulation shift?

classification 📊 stat.ML cs.LG
keywords enforcingfairnessalgorithmicdomainperformancetargetcauseddata
0
0 comments X
read the original abstract

Many instances of algorithmic bias are caused by subpopulation shifts. For example, ML models often perform worse on demographic groups that are underrepresented in the training data. In this paper, we study whether enforcing algorithmic fairness during training improves the performance of the trained model in the \emph{target domain}. On one hand, we conceive scenarios in which enforcing fairness does not improve performance in the target domain. In fact, it may even harm performance. On the other hand, we derive necessary and sufficient conditions under which enforcing algorithmic fairness leads to the Bayes model in the target domain. We also illustrate the practical implications of our theoretical results in simulations and on real data.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Mutatis Mutandis: Revisiting the Comparator in Discrimination Testing

    cs.LG 2024-05 unverdicted novelty 6.0

    The paper introduces the mutatis mutandis (MM) comparator as a causal alternative to the ceteris paribus (CP) comparator in discrimination testing, arguing that MM enables more realistic complainant-comparator pairs a...