A Statistical Framework for Evaluating the Reliability of Causal Machine Learning

• In Event: Bridging Prediction and Inference

Brief Overview

Introduces model-agnostic diagnostics to assess the credibility of causal machine learning estimators for heterogeneous treatment effects, providing tools to test model fit, identify influential cases, and detect subgroup-specific misspecification.

Authors

• Melody Huang, Yale University

• Tiffany Tang, University of Notre Dame

• Ana Maria Kenney, University of California, Irvine