Non-Discriminatory Machine Learning through Convex Fairness Criteria

By Naman Goel, Mohammad Yaghini, Boi Faltings. 2018

103 citations recorded by the source corpus.

We introduce a novel technique to achieve non-discrimination in machine learning without sacrificing convexity and probabilistic interpretation. We also propose a new notion of fairness for machine learning called the weighted proportional fairness and show that our technique satisfies this subjective fairness criterion.

Subfields: Artificial Intelligence, Safety Research, Statistics and Probability, Optimization & Training, AI Ethics, Fairness & Society

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Fred Morstatter
2. Ninareh Mehrabi
3. Faisal Mahmood
4. Aram Galstyan
5. Lingxiao Huang
6. Ming Y. Lu
7. Drew F. K. Williamson
8. L. Elisa Celis

In Safety Research

1. Nisheeth K. Vishnoi
2. L. Elisa Celis
3. Aram Galstyan
4. Fred Morstatter
5. Ninareh Mehrabi
6. Kristina Lerman
7. Lingxiao Huang
8. Vijay Keswani

In Statistics and Probability

1. Fred Morstatter
2. Ninareh Mehrabi
3. Aram Galstyan
4. Nripsuta Ani Saxena
5. L. Elisa Celis
6. Kristina Lerman
7. Nisheeth K. Vishnoi
8. Lingxiao Huang

In Optimization & Training

1. Aram Galstyan
2. L. Elisa Celis
3. Kristina Lerman
4. Nisheeth K. Vishnoi
5. Lingxiao Huang
6. Richard J. Chen
7. Vijay Keswani

In AI Ethics, Fairness & Society

1. Fred Morstatter
2. Ninareh Mehrabi
3. Lingxiao Huang
4. Aram Galstyan
5. Nisheeth K. Vishnoi
6. L. Elisa Celis
7. Kristina Lerman
8. Vijay Keswani

Citations in the corpus

• A Survey on Bias and Fairness in Machine Learning (2019, methodology): “Therefore, certain methods have been proposed [49, 65, 72, 86] that satisfy certain definitions of fairness.”
• Statistical Equity: A Fairness Classification Objective (2020, background): “Different body of work tried to incorporate the proposed definitions in different downstream tasks such as classification and regression [Menon and Williamson, 2018, Berk et al., 2017, Krasanakis et al., 2018, Agarwal, Dudik, and Wu, 2019, Goel, Yaghini, and Faltings, 2018].”
• Stable and Fair Classification (2019, methodology): “Specifically, [32] showed that the performance of GYF is comparable to ZVRG over the Adult dataset.”
• Algorithm fairness in artificial intelligence for medicine and healthcare (2023, methodology): “To mitigate the effect of confounding variables, in-processing techniques adopt a non-discrimination term within the model to penalize learning discriminative features of a protected attribute 73–75, 91, 92 .”
• Fairness in Machine Learning: A Survey (2020, unknown): “…for regularization approaches are: (1) they are often non-convex in nature or achieve convexity at the cost of probabilistic interpretation [119]; (2) not all fairness measures are equally affected by the strength of regularization parameters [26, 85]; and (3) different regularization…”
• Classification with Fairness Constraints: A Meta-Algorithm with Provable Guarantees (2018, background): “, constrained to statistical parity [29, 46, 60] or equalized odds [34, 46, 59].”