Skip to Main content Skip to Navigation
Conference papers

Improved Optimistic Algorithms for Logistic Bandits

Abstract : The generalized linear bandit framework has attracted a lot of attention in recent years by extending the well-understood linear setting and allowing to model richer reward structures. It notably covers the logistic model, widely used when rewards are binary. For logistic bandits, the frequentist regret guarantees of existing algorithms areÕ(κ √ T), where κ is a problem-dependent constant. Unfortunately, κ can be arbitrarily large as it scales exponentially with the size of the decision set. This may lead to significantly loose regret bounds and poor empirical performance. In this work, we study the logistic bandit with a focus on the prohibitive dependencies introduced by κ. We propose a new optimistic algorithm based on a finer examination of the non-linearities of the reward function. We show that it enjoys aÕ(√ T) regret with no dependency in κ, but for a second order term. Our analysis is based on a new tail-inequality for self-normalized martingales, of independent interest.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02932836
Contributor : Louis Faury <>
Submitted on : Monday, September 7, 2020 - 11:36:17 PM
Last modification on : Thursday, December 10, 2020 - 4:47:38 PM
Long-term archiving on: : Friday, December 4, 2020 - 9:12:46 PM

File

supp.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02932836, version 1

Collections

Citation

Louis Faury, Marc Abeille, Clément Calauzènes, Olivier Fercoq. Improved Optimistic Algorithms for Logistic Bandits. International Conference of Machine Learning, Jul 2020, Vienne, Austria. ⟨hal-02932836⟩

Share

Metrics

Record views

119

Files downloads

33