A Bilevel Optimization Approach for Computing Synthetic Data to Mitigate Unfairness in Collaborative Machine Learning

References

• Abadi M, Chu A, Goodfellow I, McMahan HB, Mironov I, Talwar K, Zhang L (2016) Deep learning with differential privacy. Proc. 2016 ACM SIGSAC Conf. Comput. Comm. Security (Association for Computing Machinery, New York), 308–318.Google Scholar
• Buolamwini J, Gebru T (2018) Gender shades: Intersectional accuracy disparities in commercial gender classification. Proc. 1st Conf. Fairness Accountability Transparency (PMLR, New York), 77–91.Google Scholar
• Calmon F, Wei D, Vinzamuri B, Natesan Ramamurthy K, Varshney KR (2017) Optimized pre-processing for discrimination prevention. von Luxburg U, Guyon I, Bengio S, Wallach H, Fergus R, eds. NIPS’17: Proc. 31st Internat. Conf. Neural Inform. Processing Systems (Curran Associates Inc., Red Hook, NY), 3995–4004.Google Scholar
• Caton S, Haas C (2024) Fairness in machine learning: A survey. ACM Comput. Surveys 56(7):1–38.Google Scholar
• d’Alessandro B, O’Neil C, LaGatta T (2017) Conscientious classification: A data scientist’s guide to discrimination-aware classification. Big Data 5(2):120–134.Google Scholar
• Dressel J, Farid H (2018) The accuracy, fairness, and limits of predicting recidivism. Sci. Adv. 4(1):eaao5580.Google Scholar
• Dwork C (2006) Differential privacy. Bugliesi M, Preneel B, Sassone V, Wegener I, eds. Automata Languages Programming. ICALP 2006, Lecture Notes in Computer Science, vol. 4052 (Springer, Berlin, Heidelberg), 1–12.Google Scholar
• Dwork C, Roth A (2014) The algorithmic foundations of differential privacy. Foundations Trends Theoret. Comput. Sci. 9(3–4):211–407.Google Scholar
• Ezzeldin YH, Yan S, He C, Ferrara E, Avestimehr AS (2023) FairFED: Enabling group fairness in federated learning. Proc. AAAI Conf. Artificial Intelligence 37(6):7494–7502.Google Scholar
• Fang ML, Dhami DS, Kersting K (2022) DP-CTGAN: Differentially private medical data generation using CTGANs. Michalowski M, Abidi SSR, Abidi S, eds. Artificial Intelligence Medicine. AIME 2022, Lecture Notes in Computer Science, vol. 13263 (Springer, Cham, Switzerland), 178–188.Google Scholar
• Flaticon (2023) Illustration icons. Accessed June 2, 2023, https://www.flaticon.com/.Google Scholar
• Giovannelli T, Kent GD, Vicente LN (2024) Bilevel optimization with a multi-objective lower-level problem: Risk-neutral and risk-averse formulations. Optim. Methods Software 39(4):1–23.Google Scholar
• Goetz J, Tewari A (2020) Federated learning via synthetic data. Preprint, submitted August 11, https://arxiv.org/abs/2008.04489.Google Scholar
• Goodfellow I, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozair S, Courville A, Bengio Y (2020) Generative adversarial networks. Comm. ACM 63(11):139–144.Google Scholar
• He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. Proc. IEEE Conf. Comp. Vision Pattern Recognition (IEEE Computer Society, Washington, DC), 770–778.Google Scholar
• Hsu TMH, Qi H, Brown M (2019) Measuring the effects of non-identical data distribution for federated visual classification. Preprint, submitted September 13, https://arxiv.org/abs/1909.06335.Google Scholar
• Hu S, Goetz J, Malik K, Zhan H, Liu Z, Liu Y (2022) FedSynth: Gradient compression via synthetic data in federated learning. Preprint, submitted August 4, https://arxiv.org/abs/2204.01273.Google Scholar
• Jordon J, Yoon J, Van Der Schaar M (2018) PATE-GAN: Generating synthetic data with differential privacy guarantees. Internat. Conf. Learn. Representations.Google Scholar
• Kamishima T, Akaho S, Asoh H, Sakuma J (2012) Fairness-aware classifier with prejudice remover regularizer. Flach PA, De Bie T, Cristianini N, eds. Machine Learn. Knowledge Discovery Databases. ECML PKDD 2012, Lecture Notes in Computer Science, vol. 7524 (Springer, Berlin, Heidelberg), 35–50.Google Scholar
• Kim E, Bryant D, Srikanth D, Howard A (2021) Age bias in emotion detection: An analysis of facial emotion recognition performance on young, middle-aged, and older adults. Proc. 2021 AAAI/ACM Conf. AI Ethics Soc. (Association for Computing Machinery, New York), 638–644.Google Scholar
• Kingma DP, Ba J (2015) Adam: A method for stochastic optimization. Proc. 3rd Internat. Conf. Learn. Representations (ICLR) (San Diego, CA).Google Scholar
• Laal M, Laal M (2012) Collaborative learning: What is it? Procedia-Soc. Behav. Sci. 31:491–495.Google Scholar
• Liu J, Cui P (2025) Data heterogeneity modeling for trustworthy machine learning. Proc. 31st ACM SIGKDD Conf. Knowledge Discovery Data Mining, vol. 2 (Association for Computing Machinery, New York), 6086–6095.Google Scholar
• Lohia PK, Ramamurthy KN, Bhide M, Saha D, Varshney KR, Puri R (2019) Bias mitigation post-processing for individual and group fairness. ICASSP 2019-2019 IEEE Internat. Conf. Acoustics Speech Signal Processing (ICASSP) (IEEE, Piscataway, NJ), 2847–2851.Google Scholar
• McMahan B, Ramage D (2017) Federated learning: Collaborative machine learning without centralized training data. Google AI Blog (April 6), https://research.google/blog/federated-learning-collaborative-machine-learning-without-centralized-training-data/.Google Scholar
• McMahan B, Moore E, Ramage D, Hampson S, Agüera y Arcas B (2017) Communication-efficient learning of deep networks from decentralized data. Artificial Intelligence Statist. (PMLR), 1273–1282.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 54(6):1–35.Google Scholar
• Mugunthan V, Gokul V, Kagal L, Dubnov S (2021) Bias-free FedGAN: A federated approach to generate bias-free datasets. Preprint, submitted March 17, https://arxiv.org/abs/2103.09876.Google Scholar
• Nocedal J, Wright SJ (2006) Numerical Optimization, Springer Series in Operations Research and Financial Engineering (Springer, New York).Google Scholar
• Obermeyer Z, Powers B, Vogeli C, Mullainathan S (2019) Dissecting racial bias in an algorithm used to manage the health of populations. Science 366(6464):447–453.Google Scholar
• Pan Z, Wang S, Li C, Wang H, Tang X, Zhao J (2023) FedMDFG: Federated learning with multi-gradient descent and fair guidance. Proc. AAAI Conf. Artificial Intelligence 37(8):9364–9371.Google Scholar
• Pessach D, Shmueli E (2022) A review on fairness in machine learning. ACM Comput. Surveys 55(3):1–44.Google Scholar
• Rasouli M, Sun T, Rajagopal R (2020) FedGAN: Federated generative adversarial networks for distributed data. Preprint, submitted June 12, https://arxiv.org/abs/2006.07228.Google Scholar
• Salazar T, Fernandes M, Araújo H, Abreu PH (2023) FAIR-FATE: Fair federated learning with momentum. Mikyška J, de Mulatier C, Paszynski M, Krzhizhanovskaya VV, Dongarra JJ, Sloot PM, eds. Comput. Sci. ICCS 2023, Lecture Notes in Computer Science, vol. 14073 (Springer, Cham, Switzerland), 524–538.Google Scholar
• Savard G, Gauvin J (1994) The steepest descent direction for the nonlinear bilevel programming problem. Oper. Res. Lett. 15(5):265–272.Google Scholar
• Sinha A, Malo P, Deb K (2017) A review on bilevel optimization: From classical to evolutionary approaches and applications. IEEE Trans. Evolutionary Comput. 22(2):276–295.Google Scholar
• Van der Laan P (2001) The 2001 census in the Netherlands: Integration of registers and surveys. Conf. Cathie Marsh Centre (Institut National de la Statistique et des Études Économiques, Montrouge, France), 1–24.Google Scholar
• Wang Z, Shu K, Culotta A (2021) Enhancing model robustness and fairness with causality: A regularization approach. Preprint, submitted October 3, https://arxiv.org/abs/2110.00911.Google Scholar
• Wang T, Zhu JY, Torralba A, Efros AA (2018) Dataset distillation. Preprint, submitted November 27, https://arxiv.org/abs/1811.10959.Google Scholar
• Wang J, Pal A, Yang Q, Kant K, Zhu K, Guo S (2022) Collaborative machine learning: Schemes, robustness, and privacy. IEEE Trans. Neural Networks Learn. Systems 34(12):9625–9642.Google Scholar
• Wightman LF (1998) LSAC national longitudinal bar passage study. LSAC Research Report Series, Law School Admission Council, Newtown, PA.Google Scholar
• Xie L, Lin K, Wang S, Wang F, Zhou J (2018) Differentially private generative adversarial network. Preprint, submitted February 19, https://arxiv.org/abs/1802.06739.Google Scholar
• Zafar MB, Valera I, Gomez-Rodriguez M, Gummadi KP (2019) Fairness constraints: A flexible approach for fair classification. J. Machine Learn. Res. 20(1):2737–2778.Google Scholar
• Zafar MB, Valera I, Rogriguez MG, Gummadi KP (2017) Fairness constraints: Mechanisms for fair classification. Artificial Intelligence Statist. (PMLR, New York), 962–970.Google Scholar