Reproduce “Dual Instrumental Variable Regression” 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
Krikamol Muandet, Arash Mehrjou, Si Kai Lee, Anant Raj
DualIV reformulates nonlinear instrumental-variable regression without separately estimating how treatment varies with the instrument. It jointly adjusts a causal-effect function and a function that checks residual mismatch through a min–max objective. In a simulated nonlinear demand task, its kernel version has accuracy close to established alternatives; at larger sample sizes, results are nearly indistinguishable, with no runtime advantage established.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.