TY - GEN
T1 - Using Copies to Remove Sensitive Data
T2 - 9th Iberian Conference on Pattern Recognition and Image Analysis, IbPRIA 2019
AU - Unceta Mendieta, Irene
AU - Nin, J.
AU - Pujol, Oriol
N1 - Funding Information:
This work has been partially funded by the Spanish project TIN2016-74946-P (MINECO/FEDER, UE), and by AGAUR of the Generalitat de Catalunya through the Industrial PhD grant 2017-DI-25. We gratefully acknowledge the support of BBVA Data & Analytics for sponsoring the Industrial PhD.
Funding Information:
Acknowledgment. This work has been partially funded by the Spanish project TIN2016-74946-P (MINECO/FEDER, UE), and by AGAUR of the Generalitat de Catalunya through the Industrial PhD grant 2017-DI-25. We gratefully acknowledge the support of BBVA Data & Analytics for sponsoring the Industrial PhD.
Publisher Copyright:
© 2019, Springer Nature Switzerland AG.
PY - 2019
Y1 - 2019
N2 - Ensuring classification models are fair with respect to sensitive data attributes is a crucial task when applying machine learning models to real-world problems. Particularly in company production environments, where the decision output by models may have a direct impact on individuals and predictive performance should be maintained over time. In this article, build upon [17], we propose copies as a technique to mitigate the bias of trained algorithms in circumstances where the original data is not accessible and/or the models cannot be re-trained. In particular, we explore a simple methodology to build copies that replicate the learned decision behavior in the absence of sensitive attributes. We validate this methodology in the low-sensitive problem of superhero alignment. We demonstrate that this naïve approach to bias reduction is feasible in this problem and argue that copies can be further exploited to embed models with desiderata such as fair learning.
AB - Ensuring classification models are fair with respect to sensitive data attributes is a crucial task when applying machine learning models to real-world problems. Particularly in company production environments, where the decision output by models may have a direct impact on individuals and predictive performance should be maintained over time. In this article, build upon [17], we propose copies as a technique to mitigate the bias of trained algorithms in circumstances where the original data is not accessible and/or the models cannot be re-trained. In particular, we explore a simple methodology to build copies that replicate the learned decision behavior in the absence of sensitive attributes. We validate this methodology in the low-sensitive problem of superhero alignment. We demonstrate that this naïve approach to bias reduction is feasible in this problem and argue that copies can be further exploited to embed models with desiderata such as fair learning.
KW - Bias reduction
KW - Copying classifiers
KW - Fairness
KW - Superhero alignment
UR - http://www.scopus.com/inward/record.url?scp=85076114443&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-31332-6_16
DO - 10.1007/978-3-030-31332-6_16
M3 - Conference contribution
AN - SCOPUS:85076114443
SN - 9783030313319
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 182
EP - 193
BT - Pattern Recognition and Image Analysis - 9th Iberian Conference, IbPRIA 2019, Proceedings
A2 - Morales, Aythami
A2 - Fierrez, Julian
A2 - Sánchez, José Salvador
A2 - Ribeiro, Bernardete
PB - Springer
Y2 - 1 July 2019 through 4 July 2019
ER -