ABSTRACT
A central goal of algorithmic fairness is to reduce bias in automated decision making. An unavoidable tension exists between accuracy gains obtained by using sensitive information as part of a statistical model, and any commitment to protect these characteristics. Often, due to biases present in the data, using the sensitive information in the functional form of a classifier improves classification accuracy. In this paper we show how it is possible to get the best of both worlds: optimize model accuracy and fairness without explicitly using the sensitive feature in the functional form of the model, thereby treating different individuals equally. Our method is based on two key ideas. On the one hand, we propose to use Multitask Learning (MTL), enhanced with fairness constraints, to jointly learn group specific classifiers that leverage information between sensitive groups. On the other hand, since learning group specific models might not be permitted, we propose to first predict the sensitive features by any learning method and then to use the predicted sensitive feature to train MTL with fairness constraints. This enables us to tackle fairness with a three-pronged approach, that is, by increasing accuracy on each group, enforcing measures of fairness during training, and protecting sensitive information during testing. Experimental results on two real datasets support our proposal, showing substantial improvements in both accuracy and fairness.
- J. Adebayo and L. Kagal. 2016. Iterative orthogonal feature projection for diagnosing bias in black-box models. In Conference on Fairness, Accountability, and Transparency in Machine Learning (FATML).Google Scholar
- A. Agarwal, A. Beygelzimer, M. Dud'ik, and J. Langford. 2017. A Reductions Approach to Fair Classification. In FATML.Google Scholar
- A. Agarwal, A. Beygelzimer, M. Dud'ik, J. Langford, and H. Wallach. 2018. A reductions approach to fair classification. arXiv preprint arXiv:1803.02453 (2018).Google Scholar
- D. Alabi, N. Immorlica, and A. T. Kalai. 2018. When optimizing nonlinear objectives is no harder than linear objectives. arXiv preprint arXiv:1804.04503 (2018).Google Scholar
- A. Argyriou, T. Evgeniou, and M. Pontil. 2008. Convex multi-task feature learning. Machine Learning, Vol. 73, 3 (2008), 243--272. Google ScholarDigital Library
- B. Bakker and T. Heskes. 2003. Task clustering and gating for bayesian multitask learning. Journal of Machine Learning Research, Vol. 4 (2003), 83--99. Google ScholarDigital Library
- J. Baxter. 2000. A model of inductive bias learning. Journal of artificial intelligence research, Vol. 12 (2000), 149--198. Google ScholarCross Ref
- Y. Bechavod and K. Ligett. 2018. Penalizing Unfairness in Binary Classification. arXiv preprint arXiv:1707.00044v3 (2018).Google Scholar
- R. Berk, H. Heidari, S. Jabbari, M. Joseph, M. Kearns, J. Morgenstern, S. Neel, and A. Roth. 2017. A convex framework for fair regression. arXiv preprint arXiv:1706.02409 (2017).Google Scholar
- A. Beutel, J. Chen, Z. Zhao, and E. H. Chi. 2017. Data decisions and theoretical implications when adversarially learning fair representations. In FATML.Google Scholar
- L. Breiman. 2001. Random forests. Machine learning, Vol. 45, 1 (2001), 5--32. Google ScholarDigital Library
- F. Calmon, D. Wei, B. Vinzamuri, K. Natesan Ramamurthy, and K. R. Varshney. 2017. Optimized Pre-Processing for Discrimination Prevention. In Advances in Neural Information Processing Systems (NIPS). Google ScholarDigital Library
- R. Caruana. 1997. Multitask Learning. Machine Learning, Vol. 28, 1 (1997), 41--75. Google ScholarDigital Library
- A. Chouldechova. 2017. Fair prediction with disparate impact: A study of bias in recidivism prediction instruments. Big data, Vol. 5, 2 (2017), 153--163.Google Scholar
- M. Donini, D. Martinez-Rego, M. Goodson, J. Shawe-Taylor, and M. Pontil. 2016. Distributed variance regularized multitask learning. In International Joint Conference on Neural Networks.Google Scholar
- M. Donini, L. Oneto, S. Ben-David, J. Shawe-Taylor, and M. Pontil. 2018. Empirical Risk Minimization under Fairness Constraints. In NIPS. Google ScholarDigital Library
- C. Dwork, N. Immorlica, A. T. Kalai, and M. D. M. Leiserson. 2018. Decoupled Classifiers for Group-Fair and Efficient Machine Learning. In Conference on Fairness, Accountability and Transparency (FAT).Google Scholar
- T. Evgeniou and M. Pontil. 2004. Regularized multi-task learning. In ACM SIGKDD international conference on Knowledge discovery and data mining. Google ScholarDigital Library
- M. Feldman, S. A. Friedler, J. Moeller, C. Scheidegger, and S. Venkatasubramanian. 2015. Certifying and removing disparate impact. In International Conference on Knowledge Discovery and Data Mining. Google ScholarDigital Library
- M. Hardt, E. Price, and N. Srebro. 2016. Equality of opportunity in supervised learning. In NIPS. Google ScholarDigital Library
- IBM. 2018. User-Manual CPLEX 12.7.1. IBM Software Group.Google Scholar
- F. Kamiran and T. Calders. 2009. Classifying without discriminating. In International Conference on Computer, Control and Communication.Google Scholar
- F. Kamiran and T. Calders. 2010. Classification with no discrimination by preferential sampling. In Machine Learning Conference.Google Scholar
- F. Kamiran and T. Calders. 2012. Data preprocessing techniques for classification without discrimination. Knowledge and Information Systems, Vol. 33, 1 (2012), 1--33. Google ScholarDigital Library
- T. Kamishima, S. Akaho, and J. Sakuma. 2011. Fairness-aware learning through regularization approach. In International Conference on Data Mining Workshops. Google ScholarDigital Library
- M. Kearns, S. Neel, A. Roth, and Z. S. Wu. 2017. Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness. arXiv preprint arXiv:1711.05144 (2017).Google Scholar
- A. Khosla, T. Zhou, T. Malisiewicz, A. A. Efros, and A. Torralba. 2012. Undoing the damage of dataset bias. In European Conference on Computer Vision. Google ScholarDigital Library
- A. K. Menon and R. C. Williamson. 2018. The cost of fairness in binary classification. In FAT.Google Scholar
- D. Pedreshi, S. Ruggieri, and F. Turini. 2008. Discrimination-aware data mining. In ACM SIGKDD international conference on Knowledge discovery and data mining. Google ScholarDigital Library
- A. Pérez-Suay, V. Laparra, G. Mateo-Garc'ia, J. Mu n oz-Mar'i, L. Gómez-Chova, and G. Camps-Valls. 2017. Fair Kernel Learning. In Machine Learning and Knowledge Discovery in Databases.Google Scholar
- G. Pleiss, M. Raghavan, F. Wu, J. Kleinberg, and K. Q. Weinberger. 2017. On fairness and calibration. In NIPS. Google ScholarDigital Library
- S. Shalev-Shwartz and S. Ben-David. 2014. Understanding machine learning: From theory to algorithms .Cambridge University Press. Google ScholarDigital Library
- J. Shawe-Taylor and N. Cristianini. 2004. Kernel methods for pattern analysis .Cambridge University Press. Google ScholarDigital Library
- A. J. Smola and B. Schölkopf. 2001. Learning with Kernels .MIT Press.Google Scholar
- B. Woodworth, S. Gunasekar, M. I. Ohannessian, and N. Srebro. 2017. Learning non-discriminatory predictors. In Computational Learning Theory.Google Scholar
- M. B. Zafar, I. Valera, M. Gomez Rodriguez, and K. P. Gummadi. 2017. Fairness beyond disparate treatment & disparate impact: Learning classification without disparate mistreatment. In International Conference on World Wide Web. Google ScholarDigital Library
- M. B. Zafar, I. Valera, M. Gomez Rodriguez, and K. P. Gummadi. 2017. Fairness constraints: Mechanisms for fair classification. In International Conference on Artificial Intelligence and Statistics.Google Scholar
- M. B. Zafar, I. Valera, M. Rodriguez, K. Gummadi, and A. Weller. 2017. From parity to preference-based notions of fairness in classification. In NIPS. Google ScholarDigital Library
- R. Zemel, Y. Wu, K. Swersky, T. Pitassi, and C. Dwork. 2013. Learning fair representations. In International Conference on Machine Learning. Google ScholarDigital Library
Index Terms
- Taking Advantage of Multitask Learning for Fair Classification
Recommendations
On the Power of Randomization in Fair Classification and Representation
FAccT '22: Proceedings of the 2022 ACM Conference on Fairness, Accountability, and TransparencyFair classification and fair representation learning are two important problems in supervised and unsupervised fair machine learning, respectively. Fair classification asks for a classifier that maximizes accuracy on a given data distribution subject to ...
Performance evaluation of a fair backoff algorithm for IEEE 802.11 DFWMAC
MobiHoc '02: Proceedings of the 3rd ACM international symposium on Mobile ad hoc networking & computingDue to hidden terminals and a dynamic topology, contention among stations in an ad-hoc network is not homogeneous. Some stations are at a disadvantage in opportunity of access to the shared channel and can suffer severe throughput degradation when the ...
Inter-AP coordination for fair throughput in infrastructure-based IEEE 802.11 mesh networks
IWCMC '06: Proceedings of the 2006 international conference on Wireless communications and mobile computingThis paper studies throughput fairness among different basic service sets (BSSs) in infrastructure-based IEEE 802.11 mesh networks, where inter-BSS interference is unavoidable because of the difficulty in frequency and coverage planning and the limited ...
Comments