Improving Fairness in Doubly Imbalanced Datasets

Ata Yalcin, Asli Umay Ozturk, Yigit Sever, Viktoria Pauw, Stephan Hachinger, Ismail Hakki Toroslu, Pinar Karagoz · 2026

Read Improving Fairness in Doubly Imbalanced Datasets by Ata Yalcin, Asli Umay Ozturk, Yigit Sever, Viktoria Pauw, Stephan Hachinger, Ismail Hakki Toroslu, Pinar Karagoz (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13538691.txt.

Description

Fair machine learning gets harder when a dataset is skewed twice over: few examples of the desired outcome, and few people from some protected group. Ata Yalcin and six co-authors study that double skew in a 2026 paper. They first show, through hands-on analysis, where standard debiasing methods fall short on such data.

Then they propose a method that weighs several goals at once to pick the best way to resample the data and balance both the outcome and the group counts. The aim is a dataset where fairness fixes can work without wrecking accuracy. The authors test their approach against the failures they documented.

Catalog record

CREATORAta Yalcin, Asli Umay Ozturk, Yigit Sever, Viktoria Pauw, Stephan Hachinger, Ismail Hakki Toroslu, Pinar Karagoz
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE75 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC13538691

Browse the full library · About the Tower of Babel