2 problems
- 0 votes0 replies1 view
Regret decomposition for the non-restarting doubling trick
Let be a generic bandit algorithm, and let be its non-restarting doubling-trick version, which reinitializes the algorithm at successive horizons while…
- 0 votes0 replies1 view
Exponential doubling cannot preserve the asymptotically optimal constant
Let an anytime algorithm be constructed from a non-anytime algorithm by applying a doubling trick. A geometric doubling sequence is known not to preserve the relevant problem-depen…