Reproduce “Optimal Kernel Choice for Score Function-based Causal Discovery” 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
Wenjie Wang, Biwei Huang, Feng Liu, Xinge You, Tongliang Liu, Kun Zhang, Mingming Gong
The authors introduce a score-based method for causal discovery that learns Gaussian-kernel parameters during graph search. A Gaussian process models responses as functions of proposed parents plus noise; the score favors noise independent of those parents. Using GES, the method reports best normalized structural Hamming distance in synthetic cases and higher F1 in most settings, within the model and datasets.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.