Reproduce “Learning Continuous Treatment Policy and Bipartite Embeddings for Matching with Heterogeneous Causal Effects” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Will Y. Zou, Smitha Shyam, Michael Mui, Mingshi Wang, Jan Pedersen, Zoubin Ghahramani
CTPM jointly ranks subject–treatment matches and models a continuous treatment intensity for each match, optimizing a combined outcome objective. It combines match relevance with a match-specific intensity distribution and propensity-weighted treatment-effect estimates. On Ponpare test data, it scores 0.92 c-AUC versus 0.65 for the R-learner, a reported 41% improvement; this is an estimated ranking result, not a prospective deployment effect.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.