Reproduce “Approximate Cross-validated Mean Estimates for Bayesian Hierarchical Regression 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
Amy X. Zhang, Le Bao, Changcheng Li, Michael J. Daniels
AXE makes leave-cluster-out cross-validation more manageable for Bayesian hierarchical regression models. It fits the full model once, plugs in variance-covariance posterior means, then re-estimates other fold parameters to predict held-out clusters. In Radon-subsets tests, over 99% of folds had RMSE ratios within 0.1, versus 71% for GHOST and 39% for integrated importance sampling. Its guarantees concern linear mixed models under stated conditions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.