Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
169 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Incentivized Truthful Communication for Federated Bandits (2402.04485v1)

Published 7 Feb 2024 in cs.LG and cs.GT

Abstract: To enhance the efficiency and practicality of federated bandit learning, recent advances have introduced incentives to motivate communication among clients, where a client participates only when the incentive offered by the server outweighs its participation cost. However, existing incentive mechanisms naively assume the clients are truthful: they all report their true cost and thus the higher cost one participating client claims, the more the server has to pay. Therefore, such mechanisms are vulnerable to strategic clients aiming to optimize their own utility by misreporting. To address this issue, we propose an incentive compatible (i.e., truthful) communication protocol, named Truth-FedBan, where the incentive for each participant is independent of its self-reported cost, and reporting the true cost is the only way to achieve the best utility. More importantly, Truth-FedBan still guarantees the sub-linear regret and communication cost without any overheads. In other words, the core conceptual contribution of this paper is, for the first time, demonstrating the possibility of simultaneously achieving incentive compatibility and nearly optimal regret in federated bandit learning. Extensive numerical studies further validate the effectiveness of our proposed solution.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (51)
  1. Improved algorithms for linear stochastic bandits. In NIPS, volume 11, pp.  2312–2320, 2011.
  2. Truthful auctions for pricing search keywords. In Proceedings of the 7th ACM Conference on Electronic Commerce, pp.  1–7, 2006.
  3. Truthful mechanisms for one-parameter agents. In Proceedings 42nd IEEE Symposium on Foundations of Computer Science, pp.  482–491. IEEE, 2001.
  4. Frugal path mechanisms. ACM Transactions on Algorithms (TALG), 3(1):1–22, 2007.
  5. Best arm identification in multi-armed bandits. In COLT, pp.  41–53, 2010.
  6. Finite-time analysis of the multiarmed bandit problem. Machine learning, 47(2):235–256, 2002.
  7. One for one, or all for all: Equilibria and optimality of collaboration in federated learning. In International Conference on Machine Learning, pp. 1005–1014. PMLR, 2021.
  8. Numerical analysis. Cengage learning, 2015.
  9. Edward H Clarke. Multipart pricing of public goods. Public choice, pp.  17–33, 1971.
  10. Fairness in model-sharing games. In Proceedings of the ACM Web Conference 2023, pp. 3775–3783, 2023.
  11. Differentially-private federated linear bandits. Advances in Neural Information Processing Systems, 33:6003–6014, 2020.
  12. Contextual bandits for adapting treatment in a mouse model of de novo carcinogenesis. In Machine learning for healthcare conference, pp.  67–82. PMLR, 2018.
  13. Optimal best arm identification with fixed confidence. In Conference on Learning Theory, pp.  998–1027. PMLR, 2016.
  14. Theodore Groves. Incentives in teams. Econometrica: Journal of the Econometric Society, pp. 617–631, 1973.
  15. David A Harville. Matrix Algebra From a Statistician’s Perspective. Springer Science & Business Media, 2008.
  16. A simple and provably efficient algorithm for asynchronous federated contextual linear bandits. In S. Koyejo, S. Mohamed, A. Agarwal, D. Belgrave, K. Cho, and A. Oh (eds.), Advances in Neural Information Processing Systems, volume 35, pp.  4762–4775. Curran Associates, Inc., 2022.
  17. Federated linear contextual bandits. In M. Ranzato, A. Beygelzimer, Y. Dauphin, P.S. Liang, and J. Wortman Vaughan (eds.), Advances in Neural Information Processing Systems, volume 34, pp.  27057–27068. Curran Associates, Inc., 2021.
  18. False-name-proof mechanisms for hiring a team. In Internet and Network Economics: Third International Workshop, WINE 2007, San Diego, CA, USA, December 12-14, 2007. Proceedings 3, pp.  245–256. Springer, 2007.
  19. Submodular optimization with submodular cover and submodular knapsack constraints. Advances in neural information processing systems, 26, 2013.
  20. Vcg mechanism design with unknown agent values under stochastic bandit feedback. Journal of Machine Learning Research, 24(53):1–45, 2023.
  21. Mechanisms that incentivize data sharing in federated learning. arXiv preprint arXiv:2207.04557, 2022.
  22. On distributed cooperative decision-making in multiarmed bandits. In 2016 European Control Conference (ECC), pp.  243–248. IEEE, 2016.
  23. Bandit algorithms. Cambridge University Press, 2020.
  24. An incentive mechanism for federated learning in wireless cellular networks: An auction approach. IEEE Transactions on Wireless Communications, 20(8):4874–4887, 2021.
  25. Truth revelation in approximately efficient combinatorial auctions. Journal of the ACM (JACM), 49(5):577–602, 2002.
  26. Asynchronous upper confidence bound algorithms for federated linear bandits. In International Conference on Artificial Intelligence and Statistics, pp.  6529–6553. PMLR, 2022a.
  27. Communication efficient federated learning for generalized linear bandits. In Alice H. Oh, Alekh Agarwal, Danielle Belgrave, and Kyunghyun Cho (eds.), Advances in Neural Information Processing Systems, 2022b.
  28. Communication efficient distributed learning for kernelized contextual bandits. In Alice H. Oh, Alekh Agarwal, Danielle Belgrave, and Kyunghyun Cho (eds.), Advances in Neural Information Processing Systems, 2022.
  29. Learning kernelized contextual bandits in a distributed and asynchronous environment. In The Eleventh International Conference on Learning Representations, 2023.
  30. A contextual-bandit approach to personalized news article recommendation. In Proceedings of the 19th international conference on World wide web, pp.  661–670, 2010a.
  31. Exploitation and exploration in a performance based contextual advertising system. In Proceedings of the 16th ACM SIGKDD international conference on Knowledge discovery and data mining, pp.  27–36, 2010b.
  32. Hoeffding races: Accelerating model selection search for classification and function approximation. Advances in neural information processing systems, 6, 1993.
  33. Decentralized cooperative stochastic bandits. Advances in Neural Information Processing Systems, 32, 2019.
  34. Truthful approximation mechanisms for restricted combinatorial auctions. Games and Economic Behavior, 64(2):612–631, 2008.
  35. Algorithmic mechanism design. In Proceedings of the thirty-first annual ACM symposium on Theory of computing, pp.  129–140, 1999.
  36. Jian Pei. A survey on data pricing: from economics to data science. IEEE Transactions on knowledge and Data Engineering, 34(10):4586–4608, 2020.
  37. Ariel D Procaccia. Cake cutting: Not just child’s play. Communications of the ACM, 56(7):78–87, 2013.
  38. Approximate mechanism design without money. ACM Transactions on Economics and Computation (TEAC), 1(4):1–26, 2013.
  39. Alvin E Roth. On the allocation of residents to rural hospitals: a general property of two-sided matching markets. Econometrica: Journal of the Econometric Society, pp. 425–427, 1986.
  40. Federated multi-armed bandits. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pp.  9603–9611, 2021.
  41. Decentralized multi-player multi-armed bandits with no collision information. In International Conference on Artificial Intelligence and Statistics, pp.  1519–1528. PMLR, 2020.
  42. Collaborative machine learning with incentive-aware model rewards. In International conference on machine learning, pp. 8927–8936. PMLR, 2020.
  43. Maxim Sviridenko. A note on maximizing a submodular set function subject to a knapsack constraint. Operations Research Letters, 32(1):41–43, 2004.
  44. Kunal Talwar. The price of truth: Frugality in truthful mechanisms. In Annual Symposium on Theoretical Aspects of Computer Science, pp.  608–619. Springer, 2003.
  45. Incentive mechanisms for federated learning: From economic and game theoretic perspective. IEEE transactions on cognitive communications and networking, 8(3):1566–1593, 2022.
  46. William Vickrey. Counterspeculation, auctions, and competitive sealed tenders. The Journal of finance, 16(1):8–37, 1961.
  47. Distributed bandit learning: Near-optimal regret with efficient communication. In International Conference on Learning Representations, 2020.
  48. Incentivized communication for federated bandits. In Thirty-seventh Conference on Neural Information Processing Systems, 2023.
  49. Laurence A Wolsey. An analysis of the greedy algorithm for the submodular set covering problem. Combinatorica, 2(4):385–393, 1982.
  50. Gradient driven rewards to guarantee fairness in collaborative machine learning. Advances in Neural Information Processing Systems, 34:16104–16117, 2021.
  51. Federated bandit: A gossiping approach. In Abstract Proceedings of the 2021 ACM SIGMETRICS/International Conference on Measurement and Modeling of Computer Systems, pp.  3–4, 2021.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets