Download This PaperBypassing the Monster: A Faster and Simpler Optimal Algorithm for Contextual Bandits under Realizability
Mathematics of Operations Research (to appear)
45 Pages Posted: 28 Apr 2020 Last revised: 13 Jul 2021
Date Written: March 27, 2020
Abstract
We consider the general (stochastic) contextual bandit problem under the realizability assumption, i.e., the expected reward, as a function of contexts and actions, belongs to a general function class F. We design a fast and simple algorithm that achieves the statistically optimal regret with only O(log T) calls to an offline regression oracle across all T rounds. The number of oracle calls can be further reduced to O(loglog T) if T is known in advance. Our results provide the first universal and optimal reduction from contextual bandits to offline regression, solving an important open problem in the contextual bandit literature. A direct consequence of our results is that any advances in offline regression immediately translate to contextual bandits, statistically and computationally. This leads to faster algorithms and improved regret guarantees for broader classes of contextual bandit problems.
Keywords: contextual bandit, statistical learning, offline regression, computational efficiency, reduction
Suggested Citation: Suggested Citation