# Overcoming Representation Bias in Fairness-Aware data Repair using Optimal Transport

Source: [Overcoming Representation Bias in Fairness-Aware data Repair using Optimal Transport](<https://www.sota2.com/research/paper/68fa5c5f1e209131ce9d0630>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-12T05:00:12.407Z

Authors: Abigail Langbridge, Anthony Quinn, Robert Shorten

Published: 2024-10-03T11:06:22.000Z

## Abstract

Optimal transport (OT) has an important role in transforming data distributions in a manner which engenders fairness. Typically, the OT operators are learnt from the unfair attribute-labelled data, and then used for their repair. Two significant limitations of this approach are as follows: (i) the OT operators for underrepresented subgroups are poorly learnt (i.e. they are susceptible to representation bias); and (ii) these OT repairs cannot be effected on identically distributed but out-of-sample (i.e.\\ archival) data. In this paper, we address both of these problems by adopting a Bayesian nonparametric stopping rule for learning each attribute-labelled component of the data distribution. The induced OT-optimal quantization operators can then be used to repair the archival data. We formulate a novel definition of the fair distributional target, along with quantifiers that allow us to trade fairness against damage in the transformed data. These are used to reveal excellent performance of our representation-bias-tolerant scheme in simulated and benchmark data sets.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Fairness Repair on Adult (Off-Sample)](<https://www.sota2.com/research/sota/fairness-repair-on-adult-off-sample.md>): Fairness Repair; Adult (Off-Sample); E-hat Disparity: 13.74.

- [Fairness Repair on Adult (On-Sample)](<https://www.sota2.com/research/sota/fairness-repair-on-adult-on-sample.md>): Fairness Repair; Adult (On-Sample); E-hat: 0.0041.

- [Fairness Repair on Simulated GMM dataset (Off-Sample)](<https://www.sota2.com/research/sota/fairness-repair-on-simulated-gmm-dataset-off-sample.md>): Fairness Repair; Simulated GMM dataset (Off-Sample); Log E-hat Disparity: 5.385.

- [Fairness Repair on Simulated GMM dataset (On-Sample)](<https://www.sota2.com/research/sota/fairness-repair-on-simulated-gmm-dataset-on-sample.md>): Fairness Repair; Simulated GMM dataset (On-Sample); Log Disparity E^: 13.55.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2410.02840>)

[PDF](<https://arxiv.org/pdf/2410.02840.pdf>)