Robust Optimization for Fairness with Noisy Protected Groups

Many existing fairness criteria for machine learning involve equalizing some metric across \textit{protected groups} such as race or gender. However, practitioners trying to audit or enforce such group-based criteria can easily face the problem of noisy or biased protected group information. First, we study the consequences of na{i}vely relying on noisy protected group labels: we provide an upper bound on the fairness violations on the true groups $G$ when the fairness criteria are satisfied on noisy groups $\hat{G}$. Second, we introduce two new approaches using robust optimization that, unlike the na{i}ve approach of only relying on $\hat{G}$, are guaranteed to satisfy fairness criteria on the true protected groups $G$ while minimizing a training objective. We provide theoretical guarantees that one such approach converges to an optimal feasible solution. Using two case studies, we show empirically that the robust approaches achieve better true group fairness guarantees than the na{i}ve approach.

[1]  Harikrishna Narasimhan,et al.  Optimizing Generalized Rate Metrics with Three Players , 2019, NeurIPS.

[2]  Maya R. Gupta,et al.  On Making Stochastic Classifiers Deterministic , 2019, NeurIPS.

[3]  Harikrishna Narasimhan,et al.  Pairwise Fairness for Ranking and Regression , 2019, AAAI.

[4]  Niko Kolodny,et al.  Why equality of treatment and opportunity might matter , 2019, Philosophical Studies.

[5]  D. Bandalos Measurement Theory and Applications for the Social Sciences , 2018 .

[6]  Harikrishna Narasimhan,et al.  Approximate Heavily-Constrained Learning with Lagrange Multiplier Models , 2020, NeurIPS.

[7]  Timnit Gebru,et al.  Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification , 2018, FAT.

[8]  Seth Neel,et al.  A Convex Framework for Fair Regression , 2017, ArXiv.

[9]  C. Villani Optimal Transport: Old and New , 2008 .

[10]  Percy Liang,et al.  Fairness Without Demographics in Repeated Loss Minimization , 2018, ICML.

[11]  Nathan Srebro,et al.  Learning Non-Discriminatory Predictors , 2017, COLT.

[12]  Serena Wang,et al.  Deontological Ethics By Monotonicity Shape Constraints , 2020, AISTATS.

[13]  Phebe Vayanos,et al.  Learning Optimal and Fair Decision Trees for Non-Discriminative Decision-Making , 2019, AAAI.

[14]  Anja De Waegenaere,et al.  Robust Solutions of Optimization Problems Affected by Uncertain Probabilities , 2011, Manag. Sci..

[15]  Shai Ben-David,et al.  Empirical Risk Minimization under Fairness Constraints , 2018, NeurIPS.

[16]  B. Hooks Yearning: Race, Gender, and Cultural Politics , 1990 .

[17]  Pranjal Awasthi,et al.  Equalized odds postprocessing under imperfect group information , 2019, AISTATS.

[18]  Nathan Srebro,et al.  Equality of Opportunity in Supervised Learning , 2016, NIPS.

[19]  K. Crenshaw Mapping the margins: intersectionality, identity politics, and violence against women of color , 1991 .

[20]  Constantine Caramanis,et al.  Theory and Applications of Robust Optimization , 2010, SIAM Rev..

[21]  Maya R. Gupta,et al.  Optimization with Non-Differentiable Constraints with Applications to Fairness, Recall, Churn, and Other Goals , 2018, J. Mach. Learn. Res..

[22]  Matt J. Kusner,et al.  When Worlds Collide: Integrating Different Counterfactual Assumptions in Fairness , 2017, NIPS.

[23]  Krishna P. Gummadi,et al.  Fairness Constraints: A Flexible Approach for Fair Classification , 2019, J. Mach. Learn. Res..

[24]  Emily Denton,et al.  Towards a critical race methodology in algorithmic fairness , 2019, FAT*.

[25]  Yoram Singer,et al.  Efficient projections onto the l1-ball for learning in high dimensions , 2008, ICML '08.

[26]  I-Cheng Yeh,et al.  The comparisons of data mining techniques for the predictive accuracy of probability of default of credit card clients , 2009, Expert Syst. Appl..

[27]  Esther Rolf,et al.  Delayed Impact of Fair Machine Learning , 2018, ICML.

[28]  Solon Barocas,et al.  Roles for computing in social change , 2019, FAT*.

[29]  Ben Hutchinson,et al.  50 Years of Test (Un)fairness: Lessons for Machine Learning , 2018, FAT.

[30]  Matt J. Kusner,et al.  Counterfactual Fairness , 2017, NIPS.

[31]  Nathan Srebro,et al.  Fair Learning with Private Demographic Data , 2020, ICML.

[32]  Kieran Healy,et al.  Classification situations: Life-chances in the neoliberal era , 2013 .

[33]  Maya R. Gupta,et al.  Satisfying Real-world Goals with Dataset Constraints , 2016, NIPS.

[34]  Daniel Kuhn,et al.  Data-driven distributionally robust optimization using the Wasserstein metric: performance guarantees and tractable reformulations , 2015, Mathematical Programming.

[35]  Ed H. Chi,et al.  Fairness without Demographics through Adversarially Reweighted Learning , 2020, NeurIPS.

[36]  Karthik Sridharan,et al.  Two-Player Games for Efficient Non-Convex Constrained Optimization , 2018, ALT.

[37]  Talia B. Gillis False Dreams of Algorithmic Fairness: The Case of Credit Pricing , 2020 .

[38]  John C. Duchi,et al.  Learning Models with Uniform Performance via Distributionally Robust Optimization , 2018, ArXiv.

[39]  Rediet Abebe,et al.  Fairness, Equality, and Power in Algorithmic Decision-Making , 2021, FAccT.

[40]  Suhas Vijaykumar,et al.  A Resolution in Algorithmic Fairness: Calibrated Scores for Fair Classifications , 2020, ArXiv.

[41]  Deeparnab Chakrabarty,et al.  Fair Algorithms for Clustering , 2019, NeurIPS.

[42]  Suresh Venkatasubramanian,et al.  A comparative study of fairness-enhancing interventions in machine learning , 2018, FAT.

[43]  Anthony Man-Cho So,et al.  A First-Order Algorithmic Framework for Wasserstein Distributionally Robust Logistic Regression , 2019, ArXiv.

[44]  John C. Duchi,et al.  Stochastic Gradient Methods for Distributionally Robust Optimization with f-divergences , 2016, NIPS.

[45]  Richard G. Baraniuk,et al.  Tuning Support Vector Machines for Minimax and Neyman-Pearson Classification , 2008, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[46]  Xiaojie Mao,et al.  Assessing algorithmic fairness with unobserved protected class using data combination , 2019, FAT*.

[47]  Ivar Krumpal Determinants of social desirability bias in sensitive surveys: a literature review , 2013 .

[48]  D. A. Edwards On the Kantorovich–Rubinstein theorem , 2011 .

[49]  Elad Eban,et al.  Scalable Learning of Non-Decomposable Objectives , 2016, AISTATS.

[50]  Sylvain Lamprier,et al.  Achieving Fairness with Decision Trees: An Adversarial Approach , 2020, Data Science and Engineering.

[51]  Krishna P. Gummadi,et al.  Fairness Constraints: Mechanisms for Fair Classification , 2015, AISTATS.

[52]  Abigail Z. Jacobs,et al.  Measurement and Fairness , 2019, FAccT.

[53]  Aditya Krishna Menon,et al.  Noise-tolerant fair classification , 2019, NeurIPS.

[54]  John Langford,et al.  A Reductions Approach to Fair Classification , 2018, ICML.

[55]  Toniann Pitassi,et al.  Fairness through awareness , 2011, ITCS '12.