On the fairness of disentangled representations

Locatello F, Abbati G, Rainforth T, Bauer S, Schölkopf B, Bachem O. 2019. On the fairness of disentangled representations. Advances in Neural Information Processing Systems. NeurIPS: Neural Information Processing Systems vol. 32, 14611–14624.

Download (ext.)
Conference Paper | Published | English

Scopus indexed
Author
Locatello, FrancescoISTA ; Abbati, Gabriele; Rainforth, Tom; Bauer, Stefan; Schölkopf, Bernhard; Bachem, Olivier
Department
Abstract
Recently there has been a significant interest in learning disentangled representations, as they promise increased interpretability, generalization to unseen scenarios and faster learning on downstream tasks. In this paper, we investigate the usefulness of different notions of disentanglement for improving the fairness of downstream prediction tasks based on representations. We consider the setting where the goal is to predict a target variable based on the learned representation of high-dimensional observations (such as images) that depend on both the target variable and an \emph{unobserved} sensitive variable. We show that in this setting both the optimal and empirical predictions can be unfair, even if the target variable and the sensitive variable are independent. Analyzing the representations of more than \num{12600} trained state-of-the-art disentangled models, we observe that several disentanglement scores are consistently correlated with increased fairness, suggesting that disentanglement may be a useful property to encourage fairness when sensitive variables are not observed.
Publishing Year
Date Published
2019-12-08
Proceedings Title
Advances in Neural Information Processing Systems
Volume
32
Page
14611–14624
Conference
NeurIPS: Neural Information Processing Systems
Conference Location
Vancouver, Canada
Conference Date
2019-12-08 – 2019-12-14
IST-REx-ID

Cite this

Locatello F, Abbati G, Rainforth T, Bauer S, Schölkopf B, Bachem O. On the fairness of disentangled representations. In: Advances in Neural Information Processing Systems. Vol 32. ; 2019:14611–14624.
Locatello, F., Abbati, G., Rainforth, T., Bauer, S., Schölkopf, B., & Bachem, O. (2019). On the fairness of disentangled representations. In Advances in Neural Information Processing Systems (Vol. 32, pp. 14611–14624). Vancouver, Canada.
Locatello, Francesco, Gabriele Abbati, Tom Rainforth, Stefan Bauer, Bernhard Schölkopf, and Olivier Bachem. “On the Fairness of Disentangled Representations.” In Advances in Neural Information Processing Systems, 32:14611–14624, 2019.
F. Locatello, G. Abbati, T. Rainforth, S. Bauer, B. Schölkopf, and O. Bachem, “On the fairness of disentangled representations,” in Advances in Neural Information Processing Systems, Vancouver, Canada, 2019, vol. 32, pp. 14611–14624.
Locatello F, Abbati G, Rainforth T, Bauer S, Schölkopf B, Bachem O. 2019. On the fairness of disentangled representations. Advances in Neural Information Processing Systems. NeurIPS: Neural Information Processing Systems vol. 32, 14611–14624.
Locatello, Francesco, et al. “On the Fairness of Disentangled Representations.” Advances in Neural Information Processing Systems, vol. 32, 2019, pp. 14611–14624.
All files available under the following license(s):
Copyright Statement:
This Item is protected by copyright and/or related rights. [...]

Link(s) to Main File(s)
Access Level
OA Open Access

Export

Marked Publications

Open Data ISTA Research Explorer

Sources

arXiv 1905.13662

Search this title in

Google Scholar
ISBN Search