%0 Journal Article %J ACM Conference on Health, Inference and Learning %D 2020 %T Defining Admissible Rewards for High-Confidence Policy Evaluation in Batch Reinforcement Learning %A Prasad, N. %A Engelhardt, B. %A Doshi-Velez, F. %B ACM Conference on Health, Inference and Learning %V 2 %P 1-9 %G eng