Reproduce “Causal discovery from Bayesian symbolic 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
Roger Guimerà, Kai-Hendrik Cohrs, Sergio Pedreira, Marta Sales-Pardo, Gustau Camps-Valls
causalBMS uses Bayesian symbolic regression to sample candidate equations for each time-series target, scoring directed causes by how often they appear across posterior-sampled expressions. In a nonlinear synthetic system with strong autocorrelation, it identifies the tested X-to-Y direction and lag across four noise levels, including levels where the true equation is unlearnable. These results are limited to tested systems.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.