TY - JOUR
T1 - Avoiding discrimination through causal reasoning
AU - Kilbertus, Niki
AU - Rojas-Carulla, Mateo
AU - Parascandolo, Giambattista
AU - Hardt, Moritz
AU - Janzing, Dominik
AU - Schölkopf, Bernhard
N1 - Publisher Copyright:
© 2017 Neural information processing systems foundation. All rights reserved.
PY - 2017
Y1 - 2017
N2 - Recent work on fairness in machine learning has focused on various statistical discrimination criteria and how they trade off. Most of these criteria are observational: They depend only on the joint distribution of predictor, protected attribute, features, and outcome. While convenient to work with, observational criteria have severe inherent limitations that prevent them from resolving matters of fairness conclusively. Going beyond observational criteria, we frame the problem of discrimination based on protected attributes in the language of causal reasoning. This viewpoint shifts attention from "What is the right fairness criterion?" to "What do we want to assume about our model of the causal data generating process?" Through the lens of causality, we make several contributions. First, we crisply articulate why and when observational criteria fail, thus formalizing what was before a matter of opinion. Second, our approach exposes previously ignored subtleties and why they are fundamental to the problem. Finally, we put forward natural causal non-discrimination criteria and develop algorithms that satisfy them.
AB - Recent work on fairness in machine learning has focused on various statistical discrimination criteria and how they trade off. Most of these criteria are observational: They depend only on the joint distribution of predictor, protected attribute, features, and outcome. While convenient to work with, observational criteria have severe inherent limitations that prevent them from resolving matters of fairness conclusively. Going beyond observational criteria, we frame the problem of discrimination based on protected attributes in the language of causal reasoning. This viewpoint shifts attention from "What is the right fairness criterion?" to "What do we want to assume about our model of the causal data generating process?" Through the lens of causality, we make several contributions. First, we crisply articulate why and when observational criteria fail, thus formalizing what was before a matter of opinion. Second, our approach exposes previously ignored subtleties and why they are fundamental to the problem. Finally, we put forward natural causal non-discrimination criteria and develop algorithms that satisfy them.
UR - http://www.scopus.com/inward/record.url?scp=85047020437&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:85047020437
SN - 1049-5258
VL - 2017-December
SP - 657
EP - 667
JO - Advances in Neural Information Processing Systems
JF - Advances in Neural Information Processing Systems
T2 - 31st Annual Conference on Neural Information Processing Systems, NIPS 2017
Y2 - 4 December 2017 through 9 December 2017
ER -