Abstract : Motivated by realtime website optimization, this paper is about online learning in abruptly changing environments. Two extensions of the UCBT algorithm are combined in order to handle dynamic multi-armed bandits, and specifically to cope with fast variations in the rewards. Firstly, a change point detection test based on Page-Hinkley statistics is used to overcome the limitations due to the UCBT inertia. Secondly, a controlled forgetting strategy dubbed Meta-Bandit is proposed to take care of the Exploration vs Exploitation trade-off when the PH test is triggered. Extensive empirical validation shows significant improvements compared to the baseline algorithms. The paper also investigates the sensitivity of the proposed algorithm with respect to the number of available options.
https://hal.inria.fr/inria-00164033
Contributeur : Cédric Hartland
<>
Soumis le : lundi 5 novembre 2007 - 10:18:05
Dernière modification le : mercredi 28 novembre 2018 - 15:36:02
Document(s) archivé(s) le : lundi 24 septembre 2012 - 11:15:50
