Wasserstein Robust Classification with Fairness Constraints

References

• Agarwal A, Beygelzimer A, Dudik M, Langford J, Wallach H (2018) A reductions approach to fair classification. Internat. Conf. Machine Learn. (PMLR, New York), 60–69.Google Scholar
• Aliprantis CD, Border KC (2006) Infinite Dimensional Analysis: A Hitchhiker’s Guide (Springer, Berlin).Google Scholar
• Angwin J, Larson J, Mattu S, Kirchner L (2022) Machine bias. Ethics of Data and Analytics (Auerbach Publications, Boca Raton, FL), 254–264.Google Scholar
• Baardman L, Boroujeni SB, Cohen-Hillel T, Panchamgam K, Perakis G (2023) Detecting customer trends for optimal promotion targeting. Manufacturing Service Oper. Management 25(2):448–467.Link, Google Scholar
• Bach F (2021) Learning Theory from First Principles. Draft of a book, version of, September 6, 2021, https://www.di.ens.fr/~fbach/ltfp_book.pdf.Google Scholar
• Bellamy RK, Dey K, Hind M, Hoffman SC, Houde S, Kannan K, Lohia P, et al. (2018) AI fairness 360: An extensible toolkit for detecting, understanding, and mitigating unwanted algorithmic bias. Preprint, submitted October 3, https://arxiv.org/abs/1810.01943.Google Scholar
• Berk R, Heidari H, Jabbari S, Kearns M, Roth A (2021) Fairness in criminal justice risk assessments: The state of the art. Sociol. Methods Res. 50(1):3–44.Crossref, Google Scholar
• Bertsimas D, Shtern S, Sturt B (2018) A data-driven approach to multi-stage linear optimization. Optimization Online (November 3), https://optimization-online.org/2018/11/6907/.Google Scholar
• Bertsimas D, Shtern S, Sturt B (2022) Two-stage sample robust optimization. Oper. Res. 70(1):624–640.Google Scholar
• Bird S, Dudík M, Edgar R, Horn B, Lutz R, Milan V, Sameki M, et al. (2020) Fairlearn: A toolkit for assessing and improving fairness in AI. Technical report, Microsoft, Redmond, WA.Google Scholar
• Blanchet J, Murthy K (2019) Quantifying distributional model risk via optimal transport. Math. Oper. Res. 44(2):565–600.Link, Google Scholar
• Blanchet J, Kang Y, Murthy K (2019) Robust Wasserstein profile inference and applications to machine learning. J. Appl. Probabilities 56(3):830–857.Crossref, Google Scholar
• Bose I, Mahapatra RK (2001) Business data mining—A machine learning perspective. Inform. Management 39(3):211–225.Crossref, Google Scholar
• Boyd S, Vandenberghe L (2004) Convex Optimization (Cambridge University Press, Cambridge, UK).Crossref, Google Scholar
• Brennan T, Dieterich W, Ehret B (2009) Evaluating the predictive validity of the compas risk and needs assessment system. Criminal Justice Behav. 36(1):21–40.Crossref, Google Scholar
• Calders T, Kamiran F, Pechenizkiy M (2009) Building classifiers with independency constraints. Proc. IEEE Internat. Conf. Data Mining Workshops (IEEE, Piscataway, NJ), 13–18.Google Scholar
• Chang L (2006) Applying data mining to predict college admissions yield: A case study. New Directions Institutional Res. 131:53–68.Crossref, Google Scholar
• Chapelle O, Teo C, Le Q, Smola A (2008) Tighter bounds for structured estimation. Koller D, Schuurmans D, Bengio Y, Bottou L, eds. Adv. Neural Inform. Processing Systems, vol. 21 (Curran Associates, Inc., Red Hook, NY).Google Scholar
• Chen X, Owen Z, Pixton C, Simchi-Levi D (2022) A statistical learning approach to personalization in revenue management. Management Sci. 68(3):1923–1937.Google Scholar
• Chouldechova A, Roth A (2020) A snapshot of the frontiers of fairness in machine learning. Comm. ACM 63(5):82–89.Crossref, Google Scholar
• Consumer Financial Protection Bureau (2013) CFPB and DOJ order ally to pay $80 million to consumers harmed by discriminatory auto loan pricing. Accessed October 21, 2020, https://www.consumerfinance.gov/about-us/newsroom/cfpb-and-doj-order-ally-to-pay-80-million-to-consumers-harmed-by-discriminatory-auto-loan-pricing.Google Scholar
• Corbett-Davies S, Pierson E, Feller A, Goel S, Huq A (2017) Algorithmic decision making and the cost of fairness. Proc. 23rd ACM SIGKDD Internat. Conf. Knowledge Discovery Data Mining (ACM, New York), 797–806.Google Scholar
• Dastin J (2022) Amazon scraps secret AI recruiting tool that showed bias against women. Ethics of Data and Analytics (Auerbach Publications, Boca Raton, FL), 296–299.Google Scholar
• Datta A, Tschantz MC, Datta A (2015) Automated experiments on ad privacy settings: A tale of opacity, choice, and discrimination. Proc. Privacy Enhancing Tech. 2015(1):92–112.Crossref, Google Scholar
• Donini M, Oneto L, Ben-David S, Shawe-Taylor JS, Pontil M (2018) Empirical risk minimization under fairness constraints. Bengio S, Wallach H, Larochelle H, Grauman K, Cesa-Bianchi N, Garnett R, eds. Adv. Neural Inform. Processing Systems, vol. 31 (Curran Associates, Inc., Red Hook, NY). 2791–2801.Google Scholar
• Dua D, Graff C (2017) UCI machine learning repository. Accessed October 15, 2020, http://archive.ics.uci.edu/ml.Google Scholar
• Dyer ME, Frieze AM (1988) On the complexity of computing the volume of a polyhedron. SIAM J. Comput. 17(5):967–974.Crossref, Google Scholar
• Gao R, Kleywegt AJ (2023) Distributionally robust stochastic optimization with Wasserstein distance. Math. Oper. Res. 48(2):603–655.Google Scholar
• Givens C, Shortt R (1984) A class of Wasserstein metrics for probability distributions. Michigan Math. J. 31(2):231–240.Crossref, Google Scholar
• Golrezaei N, Nazerzadeh H, Rusmevichientong P (2014) Real-time optimization of personalized assortments. Management Sci. 60(6):1532–1551.Link, Google Scholar
• Gurobi Optimization, LLC (2023) Gurobi Optimizer Reference Manual (Gurobi Optimizer, Beaverton, OR).Google Scholar
• Hardt M, Price E, Srebro N (2016) Equality of opportunity in supervised learning. Lee D, Sugiyama M, Luxburg U, Guyon I, Garnett R, eds. Adv. Neural Inform. Processing Systems, vol. 29 (Curran Associates, Inc., Red Hook, NY), 3315–3323.Google Scholar
• Hashimoto T, Srivastava M, Namkoong H, Liang P (2018) Fairness without demographics in repeated loss minimization. Proc. 35th Internat. Conf. Machine Learn., 1929–1938.Google Scholar
• Hastie T, Tibshirani R, Friedman J (2009) The Elements of Statistical Learning (Springer, Berlin).Crossref, Google Scholar
• Ho-Nguyen N, Wright SJ (2023) Adversarial classification via distributional robustness with Wasserstein ambiguity. Math. Programming 198(2):1411–1447.Google Scholar
• Jacobson T, Roszbach K (2003) Bank lending policy, credit scoring and value-at-risk. J. Bank. Finance 27(4):615–633.Crossref, Google Scholar
• Jeroslow RG (1987) Representability in mixed integer programming, I: Characterization results. Discrete Appl. Math. 17(3):223–243.Crossref, Google Scholar
• Kabakchieva D (2013) Predicting student performance by using data mining methods for classification. Cybernetics Inform. Tech. 13(1):61–72.Crossref, Google Scholar
• Kuhn D, Mohajerin Esfahani P, Nguyen VA, Shafieezadeh-Abadeh S (2019) Wasserstein distributionally robust optimization: Theory and applications in machine learning. INFORMS TutORials in Operations Research (INFORMS, Catonsville, MD), 130–166.Link, Google Scholar
• Liittschwager J, Wang C (1978) Integer programming solution of a classification problem. Management Sci. 24(14):1515–1525.Link, Google Scholar
• Lohr S (2013) Big data, trying to build better workers. The New York Times (April 21).Google Scholar
• Mak H-Y, Rong Y, Zhang J (2014) Appointment scheduling with limited distributional information. Management Sci. 61(2):316–334.Link, Google Scholar
• Mehrabi N, Morstatter F, Saxena N, Lerman K, Galstyan A (2021) A survey on bias and fairness in machine learning. ACM Comput. Surveys (CSUR) 54(6):1–35.Google Scholar
• Mišić VV, Perakis G (2020) Data analytics in operations management: A review. Manufacturing Service Oper. Management 22(1):158–169.Link, Google Scholar
• Mohajerin Esfahani P, Kuhn D (2018) Data-driven distributionally robust optimization using the Wasserstein metric: Performance guarantees and tractable reformulations. Math. Programming 171(1–2):115–166.Crossref, Google Scholar
• Monahan J, Skeem JL (2016) Risk assessment in criminal sentencing. Annu. Rev. Clinical Psych. 12:489–513.Crossref, Google Scholar
• MOSEK ApS (2024) Mosek optimizer API for python. Version 10. Accessed April 11, 2024, https://docs.mosek.com/10.1/pythonapi.pdf.Google Scholar
• Nguyen VA, Kuhn D, Mohajerin Esfahani P (2022) Distributionally robust inverse covariance estimation: The Wasserstein shrinkage estimator. Oper. Res. 70(1):490–515.Google Scholar
• Nguyen VA, Zhang F, Blanchet J, Delage E, Ye Y (2020) Distributionally robust local non-parametric conditional estimation. Larochelle H, Ranzato M, Hadsell R, Balcan MF, Lin H, eds. Adv. Neural Inform. Processing Systems, vol. 33 (Curran Associates, Inc., Red Hook, NY), 15232–15242.Google Scholar
• Obermeyer Z, Emanuel EJ (2016) Predicting the future—Big data, machine learning, and clinical medicine. New England J. Medicine 375(13):1216.Crossref, Google Scholar
• Quadrianto N, Sharmanska V (2017) Recycling privileged learning and distribution matching for fairness. Guyon I, Von Luxburg U, Bengio S, Wallach H, Fergus R, Vishwanathan S, Garnett R, eds. Adv. Neural Inform. Processing Systems, vol. 30 (Curran Associates, Inc., Red Hook, NY), 677–688.Google Scholar
• Rezaei A, Fathony R, Memarrast O, Ziebart B (2020) Fairness for robust log loss classification. Proc. AAAI Conf. Artificial Intelligence.Google Scholar
• Rudin W (1964) Principles of Mathematical Analysis, vol. 3 (McGraw-Hill, New York).Google Scholar
• Samorani M, Harris SL, Blount LG, Lu H, Santoro MA (2022) Overbooked and overlooked: Machine learning and racial bias in medical appointment scheduling. Manufacturing Service Oper. Management 24(6):2825–2842.Link, Google Scholar
• Shafieezadeh-Abadeh S, Kuhn D, Mohajerin Esfahani P (2019) Regularization via mass transportation. J. Machine Learn. Res. 20(103):1–68.Google Scholar
• Shalev-Shwartz S, Ben-David S (2014) Understanding Machine Learning: From Theory to Algorithms (Cambridge University Press, Cambridge, UK).Crossref, Google Scholar
• Shaw MJ, Gentry JA (1988) Using an expert system with inductive learning to evaluate business loans. Financial Management 17(3):45–56.Crossref, Google Scholar
• Shipp MA, Ross KN, Tamayo P, Weng AP, Kutok JL, Aguiar RCT, Gaasenbeek M, et al. (2002) Diffuse large B-cell lymphoma outcome prediction by gene-expression profiling and supervised machine learning. Nature Medicine 8(1):68–74.Crossref, Google Scholar
• Taskesen B, Blanchet J, Kuhn D, Nguyen VA (2021) A statistical test for probabilistic fairness. Proc. ACM Conf. Fairness Accountability Transparency (ACM, New York), 648–665.Google Scholar
• Taskesen B, Nguyen VA, Kuhn D, Blanchet J (2020) A distributionally robust approach to fair classification. Preprint, submitted July 18, https://arxiv.org/abs/2007.09530.Google Scholar
• Vapnik V, Vashist A (2009) A new learning paradigm: Learning using privileged information. Neural Networks 22(5–6):544–557.Crossref, Google Scholar
• Wang S, Guo W, Narasimhan H, Cotter A, Gupta M, Jordan M (2020) Robust optimization for fairness with noisy protected groups. Larochelle H, Ranzato M, Hadsell R, Balcan MF, Lin H, eds. Adv. Neural Inform. Processing Systems 33 (Curran Associates, Inc., Red Hook, NY), 5190–5203.Google Scholar
• Xie W (2020) Tractable reformulations of two-stage distributionally robust linear programs over the type ∞ Wasserstein ball. Oper. Res. Lett. 48(4):513–523.Crossref, Google Scholar
• Ye Q, Xie W (2020) Unbiased subdata selection for fair classification: A unified framework and scalable algorithms. Preprint, submitted December 24, https://arxiv.org/abs/2012.12356.Google Scholar
• Yurochkin M, Bower A, Sun Y (2020) Training individually fair ML models with sensitive subspace robustness. Proc. 8th Internat. Conf. Learn. Representations (Curran Associates, Inc., Red Hook, NY).Google Scholar
• Zafar MB, Valera I, Gomez Rodriguez M, Gummadi KP (2017) Fairness beyond disparate treatment & disparate impact: Learning classification without disparate mistreatment. Proc. 26th Internat. Conf. World Wide Web (ACM, New York), 1171–1180.Google Scholar