View Document Preview and Link Document Date: 2014-06-18 10:58:08 Open Document File Size: 295,75 KB Share Result on Facebook
City New York / Beijing / / Company EDU Satyen Kale Yahoo! Labs / MIT Press / Epoch Schedule / Intel / Microsoft / / Country COLUMBIA / China / / Facility COM Robert E. Schapire Princeton University / COM Daniel Hsu Columbia University / / IndustryTerm baseline algorithm / greedy style algorithms / coordinate descent algorithm / variant algorithm / online recommendation / epoch-greedy algorithm / online variant / measure-based algorithms / Online Cover algorithm / supervised online multiclass algorithm / main optimization algorithm / online updates / above algorithms / natural solution / contextual bandit learning algorithm / baseline algorithms / bandit algorithms / online cost-sensitive classification / important applications / fresh example using online algorithms / online oracle / computing / Online Cover / contextual bandit learning algorithms / / Organization MIT / Princeton University / IDF / Columbia University / New York / / Person Robert E. Taming / Kalai / John Langford / Alina / Daniel Hsu / Robert E. Schapire / Dean Foster / Matus Telgarsky / William R. On / Robert E. Contextual / Robert E. Predicting / / Position rt / reward rt / author / / ProvinceOrState New York / / PublishedMedium Machine Learning / Journal of Machine Learning Research / The Journal of Machine Learning Research / / Technology fresh example using online algorithms / above algorithms / Online Cover algorithm / main optimization algorithm / 2 Coordinate Descent Algorithm / baseline algorithm / measure-based algorithms / greedy style algorithms / 1 Importance-weighted LOw-Variance EpochTimed Oracleized CONtextual BANDITS algorithm / contextual bandit learning algorithm / Machine Learning / Randomized UCB algorithm / html / Robert E. Contextual bandit algorithms / epoch-greedy algorithm / optimization-based algorithms / coordinate descent algorithm / Main Results Our algorithm / contextual bandit learning algorithms / baseline algorithms / variant algorithm / / URL http / SocialTag 
Taming the Monster: A Fast and Simple Algorithm for Contextual Bandits Alekh Agarwal Microsoft Research, New York, NY ALEKHA @ MICROSOFT. COM