… bounding Simple regret for contextual bandits is … simple regret minimization in the contextual
bandit setting. We propose the Contextual-Gap algorithm, give a regret bound for the simple …

… We consider a stochastic contextual bandit model where each context s ∈ S is independently
sampled from a distribution ρ. We assume that ρ is known or that it can be approximated …

… efficient bandit algorithms for the stochastic contextual bandit setting, … regret minimization
(where we establish near-optimal minimax guarantees) versus simple regret minimization (…

… space (context) At. That is, we can think of linear bandits as single-context contextual bandits,
… For example, while linear bandits are used in recommendation systems where the set of …

… Minimax regret bounds for general policy classes The vast majority of research in contextual
bandits focuses on regret minimization. That is, for a time horizon T, the goal of the player is …

… We present a new algorithm for the contextual bandit learning problem, where the learner
repeatedly takes one of K actions in response to the observed context, and observes the …

… used to describe the contextual bandits with linear constraints … contextual bandits with
regression oracles. In this setting, the decision maker observes a context zt ∈ Z from some context …

… Designing efficient general-purpose contextual bandit algorithms that … regret guarantees can
be hopeless, alternative regret no… We propose a smooth regret notion for contextual bandits, …

… selection in contextual bandits, a simple interactive … contextual bandit learning, where a
learner must balance exploration and exploitation to make decisions online? Contextual bandit …

… In this section, we consider linear contextual bandits. Suppose that each arm a ∈ A is a
vector in Rd and |A| = K. Also, assume νs(a; µs) = N(a µs,σ2), ie, with a little abuse of notation µs(…