Strategic Usage in a Multi-Learner Setting (2401.16422v2)
Abstract: Real-world systems often involve some pool of users choosing between a set of services. With the increase in popularity of online learning algorithms, these services can now self-optimize, leveraging data collected on users to maximize some reward such as service quality. On the flipside, users may strategically choose which services to use in order to pursue their own reward functions, in the process wielding power over which services can see and use their data. Extensive prior research has been conducted on the effects of strategic users in single-service settings, with strategic behavior manifesting in the manipulation of observable features to achieve a desired classification; however, this can often be costly or unattainable for users and fails to capture the full behavior of multi-service dynamic systems. As such, we analyze a setting in which strategic users choose among several available services in order to pursue positive classifications, while services seek to minimize loss functions on their observations. We focus our analysis on realizable settings, and show that naive retraining can still lead to oscillation even if all users are observed at different times; however, if this retraining uses memory of past observations, convergent behavior can be guaranteed for certain loss function classes. We provide results obtained from synthetic and real-world data to empirically validate our theoretical findings.
- Competing bandits: The perils of exploration under competition. arXiv preprint arXiv:2007.10144, 2020.
- O. Ben-Porat and M. Tennenholtz. Best response regression. Advances in Neural Information Processing Systems, 30, 2017.
- O. Ben-Porat and M. Tennenholtz. Regression equilibrium. In Proceedings of the 2019 ACM Conference on Economics and Computation, pages 173–191, 2019.
- Strategic recourse in linear classification. ArXiv, abs/2011.00355, 2020. URL https://api.semanticscholar.org/CorpusID:226226668.
- Algorithmic censoring in dynamic learning systems. arXiv preprint arXiv:2305.09035, 2023.
- Multi-learner risk reduction under endogenous participation dynamics. arXiv preprint arXiv:2206.02667, 2022.
- Strategic classification from revealed preferences. In Proceedings of the 2018 ACM Conference on Economics and Computation, pages 55–70, 2018.
- Digital credit: A snapshot of the current landscape and open research questions. CEGA White Paper, pages 1739–76, 2017.
- Strategic classification in the dark. In International Conference on Machine Learning, pages 3672–3681. PMLR, 2021.
- Competing ai: How does competition feedback affect machine learning? In International Conference on Artificial Intelligence and Statistics, pages 1693–1701. PMLR, 2021.
- R. Gradwohl and M. Tennenholtz. Coopetition against an amazon. In International Symposium on Algorithmic Game Theory, pages 347–365. Springer, 2022.
- Strategic classification. In Proceedings of the 2016 ACM conference on innovations in theoretical computer science, pages 111–122, 2016.
- Performative power. arXiv preprint arXiv:2203.17232, 2022.
- Strategic apple tasting. arXiv preprint arXiv:2306.06250, 2023.
- Fairness without demographics in repeated loss minimization. In International Conference on Machine Learning, pages 1929–1938. PMLR, 2018.
- The disparate effects of strategic manipulation. In Proceedings of the Conference on Fairness, Accountability, and Transparency, FAT* ’19, page 259–268, New York, NY, USA, 2019. Association for Computing Machinery. ISBN 9781450361255. doi: 10.1145/3287560.3287597. URL https://doi.org/10.1145/3287560.3287597.
- Competition, alignment, and equilibria in digital marketplaces. arXiv preprint arXiv:2208.14423, 2022.
- Turning the tables: Biased, imbalanced, dynamic tabular datasets for ml evaluation, 2022.
- J. Kleinberg and M. Raghavan. How do classifiers induce agents to invest effort strategically? ACM Transactions on Economics and Computation (TEAC), 8(4):1–23, 2020.
- Competition over data: how does data purchase affect users? arXiv preprint arXiv:2201.10774, 2022.
- V. Lohweg. banknote authentication. UCI Machine Learning Repository, 2013. DOI: https://doi.org/10.24432/C55P57.
- Strategic classification is causal modeling in disguise. In International Conference on Machine Learning, pages 6917–6926. PMLR, 2020.
- The social cost of strategic classification. In Proceedings of the Conference on Fairness, Accountability, and Transparency, pages 230–239, 2019.
- Strategic representation, 2022.
- Multiplayer performative prediction: Learning in decision-dependent games. arXiv preprint arXiv:2201.03398, 2022.
- Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12:2825–2830, 2011.
- Performative prediction. In International Conference on Machine Learning, pages 7599–7609. PMLR, 2020.
- G. Piliouras and F.-Y. Yu. Multi-agent performative prediction: From global stability and optimality to chaos. arXiv preprint arXiv:2201.10483, 2022.
- K. Wood and E. Dall’Anese. Stochastic saddle point problems with decision-dependent distributions. arXiv preprint arXiv:2201.02313, 2022.
- Understanding deep learning requires rethinking generalization, 2017.
- Group retention when using machine learning in sequential decision making: the interplay between user dynamics and fairness. Advances in Neural Information Processing Systems, 32, 2019.
- Who leads and who follows in strategic classification? Advances in Neural Information Processing Systems, 34:15257–15269, 2021.
- Eliot Shekhtman (3 papers)
- Sarah Dean (37 papers)