Policy Evaluation with Latent Confounders via
Optimal Balance
Andrew Bennett Nathan Kallus
Cornell University Cornell University
awb222@cornell.edu kallus@cornell.edu
Abstract
Evaluating novel contextual bandit policies using logged data is crucial in appli-
cations where exploration is costly, such as medicine. But it usually relies on the
assumption of no unobserv ...


雷达卡


京公网安备 11010802022788号







