NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:1879
Title:PC-Fairness: A Unified Framework for Measuring Causality-based Fairness


		
While reviewers agree that the proposed framework is an important addition in unifying notions of counterfactual fairness, reviewers point out limitations that could reduce its practical impact. While reviewers agree that the authors address this to some extent, more discussion of the limitations would be good to add for the camera ready version (along with various revisions suggested by the reviewers including re-organizing the paper to make it clearer). I agree with R2 in recommending that the authors change the abbreviation of PC effect.