Jul 2026
Bounding the Causal Impact of ML-assisted Decision-Making via Counterfactual Correctness
This work presents a partial-identification approach to using prior RCT data to construct bounds on the causal effect of a new model, and leverages assumptions relating fine-grained predictive accuracy to downstream outcomes.
Jonathan Zhang, Erik Skalnes, Jacob M. Chen et al.
· arXiv.org · 0 citations