主 题: Sequential Allocations with Covariates
报告人: Prof. Yuhong Yang (University of Minnesota)
时 间: 2006-07-11 下午 2:00 - 3:30
地 点: 理科一号楼 1570
Sequential clinical trials require proper allocations of the treatments to
the patients one after another. This is just the traditional interesting
bandit problem, a version of which is as follows. Suppose that there are K
(K>1) possibly biased coins available for play. At each time, you can choose
one and only one coin to flip, and win a dollar if you get a head and receive
nothing otherwise. Your goal is to obtain as much money as possible at the
end of the game. Besides clinical trials, bandit problems have applications
in e.g., scheduling and automated problem-solving in machine learning. In the
literature, covariates, while available in most applications, are rarely
considered. In this talk, I will present a non-parametric approach with a
randomization technique for effectively utilizing the information in the
covariates.