Towards Estimating Bounds on the Effect of Policies under Unobserved Confounding

Neural Information Processing Systems 

As many practical fields transition to provide personalized decisions, data is increasingly relevant to support the evaluation of candidate plans and policies ( e.g.,