Timezone: »

 
Poster
A KL-LUCB algorithm for Large-Scale Crowdsourcing
Ervin Tanczos · Robert Nowak · Bob Mankoff

Wed Dec 06 06:30 PM -- 10:30 PM (PST) @ Pacific Ballroom #28
in Posters Wed »

This paper focuses on best-arm identification in multi-armed bandits with bounded rewards. We develop an algorithm that is a fusion of lil-UCB and KL-LUCB, offering the best qualities of the two algorithms in one method. This is achieved by proving a novel anytime confidence bound for the mean of bounded distributions, which is the analogue of the LIL-type bounds recently developed for sub-Gaussian distributions. We corroborate our theoretical results with numerical experiments based on the New Yorker Cartoon Caption Contest.

Keywords: Bandit Algorithms
[ Paper] 

Author Information

Ervin Tanczos (University of Wisconsin - Madison)
Robert Nowak (University of Wisconsion-Madison)
Bob Mankoff (Former Cartoon Editor of The New Yorker)

More from the Same Authors