TY  - RPRT
TI  - KL-UCB-switch: optimal regret bounds for stochastic bandits from both a distribution-dependent and a distribution-free viewpoints
AU  - Aurélien Garivier
AU  - Hédi Hadiji
AU  - Pierre Menard
AU  - Gilles Stoltz
PY  - 2022
UR  - https://arxiv.org/abs/1805.05071
ID  - 1805.05071
ER  -