Reproduce “Instrument Space Selection for Kernel Maximum Moment Restriction” 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
Rui Zhang, Krikamol Muandet, Bernhard Schölkopf, Masaaki Imaizumi
The authors introduce KMMR instrument-space selection for instrumental-variable regression, choosing the least complex reproducing kernel Hilbert space that identifies model parameters. Their least identification selection criterion combines an identification test with a complexity-penalized risk. In a weak-instrument setting, mean squared error was 0.024 ± 0.010, versus 0.214 ± 0.019 for the median heuristic; under strong instruments, performance was nearly tied.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.