Tackling algorithmic bias in neural-network classifiers using Wasserstein-2 regularization

DOI10.1007/s10851-022-01090-2OpenAlexW4288262450WikidataQ123246212 ScholiaQ123246212MaRDI QIDQ2103876

Laurent Risser, Quentin Vincenot, Alberto González Sanz, Jean-Michel Loubes

Publication date: 9 December 2022

Published in: Journal of Mathematical Imaging and Vision (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1908.05783

zbMATH Keywords

regularizationshape recognitionimage classificationalgorithmic biasneural-networks

Mathematics Subject Classification ID

Computer science (68-XX) Information and communication theory, circuits (94-XX)

Related Items

An Improved Central Limit Theorem and Fast Convergence Rates for Entropic Transportation Costs

Uses Software

• ImageNet
• AdaGrad
• Adam
• AlexNet

Cites Work

• Empirical optimal transport on countable metric spaces: distributional limits and statistical applications
• Convergence of a Newton algorithm for semi-discrete optimal transport
• Optimal transport for applied mathematicians. Calculus of variations, PDEs, and modeling
• Central limit theorems for empirical transportation cost in general dimension
• An algorithm for removing sensitive information: application to race-independent recidivism prediction
• Inference for Empirical Wasserstein Distances on Finite Spaces
• Optimization Methods for Large-Scale Machine Learning
• A central limit theorem for Lp transportation cost on the real line with application to fairness assessment in machine learning
• Anchor Regression: Heterogeneous Data Meet Causality
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item