Algorithms using Follow-The-Perturbed-Leader and Counting Asleep Times deliver regret bounds for full-information, bandit, and restricted-information feedback in online learning with stochastically available composite actions, improving guarantees for the sleeping bandit problem.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
Introduces implicit exploration to achieve near-optimal regret in bandit problems with side observations and a related combinatorial setting without prior knowledge of the feedback structure.
citing papers explorer
-
Online combinatorial optimization with stochastic decision sets and adversarial losses
Algorithms using Follow-The-Perturbed-Leader and Counting Asleep Times deliver regret bounds for full-information, bandit, and restricted-information feedback in online learning with stochastically available composite actions, improving guarantees for the sleeping bandit problem.
-
Efficient learning by implicit exploration in bandit problems with side observations
Introduces implicit exploration to achieve near-optimal regret in bandit problems with side observations and a related combinatorial setting without prior knowledge of the feedback structure.