Get Started
Home
Topics
Search
Library
Research questionHow can we recover a target site's full treated-outcome distribution from controls alone under cross-site heterogeneity?The source site provides treated and control observations, while the target site provides only controls. Differences in measurement systems, observed features, population composition, and latent context make the target treated distribution difficult to infer from source data.
Health
Machine Learning
Research Paper
Statistical Machine Learning
Latest papersRecent research connected to this question, newest first.Distributional Treatment Effect Transportability across Heterogeneous SitesThe setting assumes treated and control units are observed at the source site and only controls at the target site. The framework models cross-site differences through a transformation learned by comparing source and target controls, then uses it to construct a synthetic target treated sample. The input reports theoretical convergence results, simulations, and a patient-derived xenograft application; it does not establish applicability beyond these settings.research paper · Sep 3, 2026
Related questions
How can extreme quantile treatment effects be estimated under heavy tails while respecting location invariance?How can we reliably select a heterogeneous treatment effect estimator without observing treatment effects?How can sequential treatment allocation learn unknown outcome variances while preserving efficient ATE inference?How can we estimate counterfactual outcomes with valid uncertainty as treatment patterns shift?