Non-Asymptotic Analysis of a UCB-based Top Two Algorithm
share
A Top Two sampling rule for bandit identification is a method which selects the next arm to sample from among two candidate arms, a leader and a challenger. Due to their simplicity and good empirical performance, they have received increased attention in recent years. For fixed-confidence best arm identification, theoretical guarantees for Top Two methods have only been obtained in the asymptotic regime, when the error level vanishes. We derive the first non-asymptotic upper bound on the expected sample complexity of a Top Two algorithm holding for any error level. Our analysis highlights sufficient properties for a regret minimization algorithm to be used as leader. They are satisfied by the UCB algorithm and our proposed UCB-based Top Two algorithm enjoys simultaneously non-asymptotic guarantees and competitive empirical performance.
READ FULL TEXT