Fair machine learning through constrained stochastic optimization and an ϵ-constraint method

A strategy for fair supervised learning is proposed. It involves formulating an optimization problem to minimize loss subject to a prescribed bound on a measure of unfairness (e.g., disparate impact). It can be embedded within an ϵ -constraint method for multiobjective optimization, allowing one to...

Full description

Saved in:
Bibliographic Details
Published inOptimization letters Vol. 18; no. 9; pp. 1975 - 1991
Main Authors Curtis, Frank E., Liu, Suyun, Robinson, Daniel P.
Format Journal Article
LanguageEnglish
Published Berlin/Heidelberg Springer Berlin Heidelberg 01.12.2024
Subjects
Computational Intelligence
Mathematics
Mathematics and Statistics
Numerical and Computational Physics
Operations Research/Decision Theory
Optimization
Original Paper
Simulation
Stochastic multiobjective optimization
Fair machine learning
Online AccessGet full text

Cover

More Information
Summary:A strategy for fair supervised learning is proposed. It involves formulating an optimization problem to minimize loss subject to a prescribed bound on a measure of unfairness (e.g., disparate impact). It can be embedded within an ϵ -constraint method for multiobjective optimization, allowing one to produce a Pareto front for minimizing loss and unfairness. A stochastic optimization algorithm, designed to be scalable for large data settings, is proposed for solving the arising constrained optimization problems. Numerical experiments on problems pertaining to predicting recidivism and income provide evidence that the strategy can be effective for large-scale fair learning.
ISSN:1862-4472
1862-4480
DOI:10.1007/s11590-023-02024-6