Abstract
We study online regret minimization algorithms in a bicriteria setting, examining not only the standard notion of regret to the best expert, but also the regret to the average of all experts, the regret to any fixed mixture of experts, and the regret to the worst expert. This study leads both to new understanding of the limitations of existing no-regret algorithms, and to new algorithms with novel performance guarantees. More specifically, we show that any algorithm that achieves only \(O(\sqrt{T})\) cumulative regret to the best expert on a sequence of T trials must, in the worst case, suffer regret \(\Omega(\sqrt{T})\) to the average, and that for a wide class of update rules that includes many existing no-regret algorithms (such as Exponential Weights and Follow the Perturbed Leader), the product of the regret to the best and the regret to the average is Ω(T). We then describe and analyze a new multi-phase algorithm, which achieves cumulative regret only \(O(\sqrt{T}\log T)\) to the best expert and has only constant regret to any fixed distribution over experts (that is, with no dependence on either T or the number of experts N). The key to the new algorithm is the gradual increase in the “aggressiveness” of updates in response to observed divergences in expert performances.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Littlestone, N., Warmuth, M.K.: The weighted majority algorithm. Information and Computation 108(2), 212–261 (1994)
Freund, Y.: Predicting a binary sequence almost as well as the optimal biased coin. Information and Computation 182(2), 73–94 (2003)
Kalai, A., Vempala, S.: Efficient algorithms for on-line optimization. Journal of Computer and System Sciences 71(3), 291–307 (2005)
Cesa-Bianchi, N., Mansour, Y., Stoltz, G.: Improved second-order bounds for prediction with expert advice. In: COLT, pp. 217–232 (2005)
Auer, P., Cesa-Bianchi, N., Gentile, C.: Adaptive and self-confident on-line learning algorithms. Journal of Computer and System Sciences 64, 48–75 (2002)
Vovk, V.: A game of prediction with expert advice. Journal of Computer and System Sciences 56(2), 153–173 (1998)
Cesa-Bianchi, N., Lugosi, G.: Prediction, learning, and games. Cambridge University Press, Cambridge (2006)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2007 Springer Berlin Heidelberg
About this paper
Cite this paper
Even-Dar, E., Kearns, M., Mansour, Y., Wortman, J. (2007). Regret to the Best vs. Regret to the Average. In: Bshouty, N.H., Gentile, C. (eds) Learning Theory. COLT 2007. Lecture Notes in Computer Science(), vol 4539. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-72927-3_18
Download citation
DOI: https://doi.org/10.1007/978-3-540-72927-3_18
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-72925-9
Online ISBN: 978-3-540-72927-3
eBook Packages: Computer ScienceComputer Science (R0)