Timezone: »

Nonparametric Contextual Bandits in Metric Spaces with Unknown Metric
Nirandika Wanigasekara · Christina Yu

Wed Dec 11 05:00 PM -- 07:00 PM (PST) @ East Exhibition Hall B + C #7
Consider a nonparametric contextual multi-arm bandit problem where each arm $a \in [K]$ is associated to a nonparametric reward function $f_a: [0,1] \to \mathbb{R}$ mapping from contexts to the expected reward. Suppose that there is a large set of arms, yet there is a simple but unknown structure amongst the arm reward functions, e.g. finite types or smooth with respect to an unknown metric space. We present a novel algorithm which learns data-driven similarities amongst the arms, in order to implement adaptive partitioning of the context-arm space for more efficient learning. We provide regret bounds along with simulations that highlight the algorithm's dependence on the local geometry of the reward functions.

Author Information

Nirandika Wanigasekara (National University of Singapore)
Christina Yu (Cornell University)

More from the Same Authors