Efficient Change-Point Detection for Tackling Piecewise-Stationary Bandits - Archive ouverte HAL Access content directly
Journal Articles Journal of Machine Learning Research Year : 2022

Efficient Change-Point Detection for Tackling Piecewise-Stationary Bandits

(1) , (2) , (2) , (2, 3)
Lilian Besson
Emilie Kaufmann
Odalric-Ambrym Maillard
Julien Seznec
  • Function : Author
  • PersonId : 1084851


We introduce GLR-klUCB, a novel algorithm for the piecewise iid non-stationary bandit problem with bounded rewards. This algorithm combines an efficient bandit algorithm, kl-UCB, with an efficient, parameter-free, changepoint detector, the Bernoulli Generalized Likelihood Ratio Test, for which we provide new theoretical guarantees of independent interest. Unlike previous non-stationary bandit algorithms using a change-point detector, GLR-klUCB does not need to be calibrated based on prior knowledge on the arms' means. We prove that this algorithm can attain a $O(\sqrt{TA \Upsilon_T\log(T)})$ regret in $T$ rounds on some ``easy'' instances, where A is the number of arms and $\Upsilon_T$ the number of change-points, without prior knowledge of $\Upsilon_T$. In contrast with recently proposed algorithms that are agnostic to $\Upsilon_T$, we perform a numerical study showing that GLR-klUCB is also very efficient in practice, beyond easy instances.
Fichier principal
Vignette du fichier
BKMS22 (1).pdf (668.02 Ko) Télécharger le fichier
Origin : Publisher files allowed on an open archive

Dates and versions

hal-02006471 , version 1 (04-02-2019)
hal-02006471 , version 2 (08-12-2020)
hal-02006471 , version 3 (01-08-2022)


Attribution - NonCommercial - ShareAlike - CC BY 4.0



Lilian Besson, Emilie Kaufmann, Odalric-Ambrym Maillard, Julien Seznec. Efficient Change-Point Detection for Tackling Piecewise-Stationary Bandits. Journal of Machine Learning Research, 2022. ⟨hal-02006471v3⟩
486 View
751 Download



Gmail Facebook Twitter LinkedIn More