Reproduce “Skewness-Robust Causal Discovery in Location-Scale Noise Models” 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
Daniel Klippert, Alexander Marx
SKEWD performs bivariate causal discovery with location-scale noise models whose noise is skew-normal rather than necessarily Gaussian. It estimates location and scale functions, refines them with expectation conditional maximization, and scores directions by likelihood or HSIC residual-independence tests. On synthetic LSs(1.750) pairs, SKEWD-LL reached 100% accuracy, versus 66% for LOCI-LL and 81% for ROCHE, under that benchmark.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.