Abstract
We consider federated learning settings with independent, self-interested participants. As all contributions are made privately, participants may be tempted to free-ride and provide redundant or low-quality data while still enjoying the benefits of the FL model. In Federated Learning, this is especially harmful as low-quality data can degrade the quality of the FL model.
Free-riding can be countered by giving incentives to participants to provide truthful data. While there are game-theoretic schemes for rewarding truthful data, they do not take into account redundancy of data with previous contributions. This creates arbitrage opportunities where participants can gain rewards for redundant data, and the federation may be forced to pay out more incentives than justified by the value of the FL model.
We show how a scheme based on influence can both guarantee that the incentive budget is bounded in proportion to the value of the FL model, and that truthfully reporting data is the dominant strategy of the participants. We show that under reasonable conditions, this result holds even when the testing data is provided by participants.
Supported by EPFL.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
It is straightforward to extend the results in this chapter to a setting where increased effort results in increased quality, but this would require to characterize the exact relation which depends on the application.
References
Yang, Q., Liu, Y., Chen, T., Tong, Y.: Federated machine learning: concept and applications. ACM Trans. Intell. Syst. Technol. 10(2), Article 12 (2019). 19 p. https://doi.org/10.1145/3298981
Liu, Y., Wei, J.: Incentives for federated learning: a hypothesis elicitation approach. In: ICML Workshop on Incentives in Machine Learning (2020)
Yu, H., et al.: A sustainable incentive scheme for federated learning. IEEE Intell. Syst. 35(4) (2020). https://doi.org/10.1109/MIS.2020.2987774
Richardson, A., Filos-Ratsikas, A., Rokvic, L., Faltings, B.: Privately computing influence in regression models. In: AAAI 2020 Workshop on Privacy-Preserving Artificial Intelligence (2020)
Cai, Y., Daskalakis, C., Papadimitriou, C.: Optimum statistical estimation with strategic data sources. In: Grünwald, P., Hazan, E., Kale, S., (eds.) Proceedings of The 28th Conference on Learning Theory, Proceedings of Machine Learning Research, Paris, France, vol. 40, pp. 280–296. PMLR (2015)
Caragiannis, I., Procaccia, A., Shah, N.: Truthful univariate estimators. In: International Conference on Machine Learning, pp. 127–135 (2016)
Chen, Y., Immorlica, N., Lucier, B., Syrgkanis, V., Ziani, J.: Optimal data acquisition for statistical estimation. In: Proceedings of the 2018 ACM Conference on Economics and Computation, pp. 27–44. ACM (2018)
Chen, Y., Podimata, C., Procaccia, A.D., Shah, N.: Strategyproof linear regression in high dimensions. In: Proceedings of the 2018 ACM Conference on Economics and Computation, pp. 9–26. ACM (2018)
Cook, R.D., Weisberg, S.: Characterizations of an empirical influence function for detecting influential cases in regression. Technometrics 22(4), 495–508 (1980)
Dwork, C.: Differential privacy: a survey of results. In: Agrawal, M., Du, D., Duan, Z., Li, A. (eds.) TAMC 2008. LNCS, vol. 4978, pp. 1–19. Springer, Heidelberg (2008). https://doi.org/10.1007/978-3-540-79228-4_1
Faltings, B., Radanovic, G.: Game theory for data science: eliciting truthful information. In: Synthesis Lectures on Artificial Intelligence and Machine Learning, vol. 11, no. 2, pp. 1–151 (2017)
Faltings, B., Jurca, R., Radanovic, G.: Peer truth serum: incentives for crowdsourcing measurements and opinions. CoRR abs/1704.05269 (2017)
Jia, R., et al.: Towards efficient data valuation based on the Shapley value. In: Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics (AISTATS) (2019)
Koh, P.W., Liang, P.: Understanding black-box predictions via influence functions. In: Precup, D., Teh, Y.W. (eds.) Proceedings of the 34th International Conference on Machine Learning, Proceedings of Machine Learning Research, International Convention Centre, Sydney, Australia, vol. 70, pp. 1885–1894. PMLR (2017)
Loog, M., Viering, T., Mey, A.: Minimizers of the empirical risk and risk monotonicity. In: Advances in Neural Information Processing Systems, pp. 7476–7485 (2019)
Radanovic, G., Faltings, B., Jurca, R.: Incentives for effort in crowdsourcing using the peer truth serum. ACM Trans. Intell. Syst. Technol. (TIST) 7(4), 48 (2016)
Shnayder, V., Agarwal, A., Frongillo, R., Parkes, D.C.: Informed truthfulness in multi-task peer prediction. In: Proceedings of the 2016 ACM Conference on Economics and Computation, pp. 179–196 (2016)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Richardson, A., Filos-Ratsikas, A., Faltings, B. (2020). Budget-Bounded Incentives for Federated Learning. In: Yang, Q., Fan, L., Yu, H. (eds) Federated Learning. Lecture Notes in Computer Science(), vol 12500. Springer, Cham. https://doi.org/10.1007/978-3-030-63076-8_13
Download citation
DOI: https://doi.org/10.1007/978-3-030-63076-8_13
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-63075-1
Online ISBN: 978-3-030-63076-8
eBook Packages: Computer ScienceComputer Science (R0)