,
Perturbed-History Exploration in Stochastic Multi-Armed Bandits
IJCAI, (июня 2019)
Аннотация
We propose an online algorithm for cumulative regret minimization in a stochastic multi-armed ban-dit. The algorithm adds O(t) i.i.d. pseudo-rewards to its history in round t and then pulls the arm with the highest average reward in its perturbed history. Therefore, we call it perturbed-history exploration (PHE). The pseudo-rewards are carefully designed to offset potentially underestimated mean rewards of arms with a high probability. We derive near-optimal gap-dependent and gap-free bounds on then-round regret of PHE. The key step in our analysis is a novel argument that shows that randomized Bernoulli rewards lead to optimism. Finally, we empirically evaluate PHE and show that it is competitive with state-of-the-art baselines.
