Statistically Efficient, Polynomial Time Algorithms for Combinatorial Semi Bandits - Archive ouverte HAL Access content directly
Journal Articles Proceedings of the ACM on Measurement and Analysis of Computing Systems Year : 2021

Statistically Efficient, Polynomial Time Algorithms for Combinatorial Semi Bandits

Algorithmes statistiquement efficaces et en temps polynomial pour les semi-bandits combinatoires

Thibaut Cuvelier
Richard Combes
Eric Gourdin
  • Function : Author
  • PersonId : 952670

Abstract

We consider combinatorial semi-bandits over a set of arms X ⊂ {0, 1} d where rewards are uncorrelated across items. For this problem, the algorithm ESCB yields the smallest known regret bound R(T) = O d(ln m)² (ln T) ∆ min , but it has computational complexity O(|X|) which is typically exponential in d, and cannot be used in large dimensions. We propose the first algorithm which is both computationally and statistically efficient for this problem with regret R(T) = O d(ln m) 2 (ln T) ∆ min and computational complexity O(T poly(d)). Our approach involves carefully designing an approximate version of ESCB with the same regret guarantees, showing that this approximate algorithm can be implemented in time O(T poly(d)) by repeatedly maximizing a linear function over X subject to a linear budget constraint, and showing how to solve this maximization problems efficiently.
Fichier principal
Vignette du fichier
2002.07258.pdf (621.61 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03162127 , version 1 (08-03-2021)

Identifiers

Cite

Thibaut Cuvelier, Richard Combes, Eric Gourdin. Statistically Efficient, Polynomial Time Algorithms for Combinatorial Semi Bandits. Proceedings of the ACM on Measurement and Analysis of Computing Systems , 2021, 5 (9), pp.1-31. ⟨10.1145/3447387⟩. ⟨hal-03162127⟩
41 View
37 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More