Reproduce “Gaussian Process Structural Equation Models with Latent Variables” 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
Ricardo Silva, Robert B. Gramacy
GPSEM-LV introduces a Bayesian latent-variable structural equation model that learns flexible Gaussian-process links between hidden factors while treating observed indicators as noisy linear measurements. It approximates each link with sampled pseudo-inputs, rather than optimized locations, and assumes a known graph. In a 150-case simulation, its estimates recover a noisy quadratic relationship; this demonstrates recovery in that setup, not general superiority.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.