Fetching the paper…

Reading the bibliography…

2018

Gradient Reversal Against Discrimination

Raff, Edward, Sylvester, Jared

Original
PDF
Post on X

Understand

No methods currently exist for making arbitrary neural networks fair.

In this work we introduce GRAD, a new and simplified method to producing fair neural networks that can be used for auto-encoding fair representations or directly with predictive networks.
It is easy to implement and add to existing architectures, has only one (insensitive) hyper-parameter, and provides improved individual and group fairness.
We use the flexibility of GRAD to demonstrate multi-attribute protection.

Gradient Reversal Against Discrimination · Around

Built on

Discrimination-aware Data Mining
Pedreshi, Dino, Ruggieri, Salvatore, and Turini, Franco · 2008
Earlier work this paper cites.
Learning Fair Representations
Zemel, Rich, Wu, Yu, Swersky, Kevin, Pitassi, Toni, and Dwork, Cynthia · 2013
Earlier work this paper cites.
A multidisciplinary survey on discrimination analysis
Romei, Andrea and Ruggieri, Salvatore · 2014
Earlier work this paper cites.
Chainer: a Next-Generation Open Source Framework for Deep Learning
Tokui, Seiya, Oono, Kenta, Hido, Shohei, and Clayton, Justin · 2015
Earlier work this paper cites.

Similar

Censoring Representations with an Adversary
Original
Edwards, Harrison and Storkey, Amos · 2016
Cited alongside, same era.
Domain-adversarial Training of Neural Networks
Ganin, Yaroslav, Ustinova, Evgeniya, Ajakan, Hana, Germain, Pascal, Larochelle, Hugo, Laviolette, François, Marchand, Mario, and Lempitsky, Victor · 2016
Cited alongside, same era.
Robust Text Classification in the Presence of Confounding Bias
Landeiro, Virgile and Culotta, Aron · 2016
Cited alongside, same era.
The Variational Fair Autoencoder
Original
Louizos, Christos, Swersky, Kevin, Li, Yujia, Welling, Max, and Zemel, Richard · 2016
Cited alongside, same era.

Then

Learning Fair Classifiers: A Regularization-Inspired Approach
Original
Bechavod, Yahav and Ligett, Katrina · 2017
Later among the works it cites.
Decoupled classifiers for fair and efficient machine learning
Original
Dwork, Cynthia, Immorlica, Nicole, Kalai, Adam Tauman, and Leiserson, Max · 2017
Later among the works it cites.
Fairness constraints: Mechanisms for fair classification
Zafar, Muhammad Bilal, Valera, Isabel, Rogriguez, Manuel Gomez, and Gummadi, Krishna P · 2017
Later among the works it cites.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…