Heterogeneous Treatment Effect Estimation

If you’ve ever shipped a machine learning model that nailed validation metrics but failed in production, you’ve felt the gap between prediction and decision. Causality is how we bridge that gap—by asking “what would happen if we change X?” instead of merely “what tends to happen when X is high?” It’s the difference between correlation…