Bias-Robust Bayesian Optimization via Dueling Bandits
Johannes Kirschner 1 Andreas Krause 1
Abstract We study a setting where the learner’s objective is to max-
imize an unknown function f : X → R with additive
We consider Bayesian optimization in settings confounded feedback,
where observations can be adversarially biased,
yt = f (xt ) + bt + t , ( ...


雷达卡




京公网安备 11010802022788号







