On Explore-Then-Commit Strategies
This addresses a fundamental limitation in bandit algorithm design for researchers and practitioners, though it is incremental as it builds on existing deviation inequalities.
The paper demonstrates that explore-then-commit strategies for two-armed Gaussian bandits are inherently suboptimal for minimizing regret, regardless of whether the arm means are known, and provides sequential strategies with asymptotically optimal and minimax order-optimal finite-time regret guarantees.
We study the problem of minimising regret in two-armed bandit problems with Gaussian rewards. Our objective is to use this simple setting to illustrate that strategies based on an exploration phase (up to a stopping time) followed by exploitation are necessarily suboptimal. The results hold regardless of whether or not the difference in means between the two arms is known. Besides the main message, we also refine existing deviation inequalities, which allow us to design fully sequential strategies with finite-time regret guarantees that are (a) asymptotically optimal as the horizon grows and (b) order-optimal in the minimax sense. Furthermore we provide empirical evidence that the theory also holds in practice and discuss extensions to non-gaussian and multiple-armed case.