We study the greedy (exploitation-only) algorithm in bandit problems with a known reward structure. We allow arbitrary finite reward structures, while prior work focused on a few specific ones. We ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback