Reproduce “Large-Scale Bayesian Optimisation Using Uncertainty-Calibrated Product-of-Experts Gaussian Process Models with Trust Regions” 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
Yean Hoon Ong
BO-pro-c-tr combines a calibrated product-of-experts Gaussian-process surrogate with flexible trust-region links, allowing local experts to support multiple regions while sharing observations. Across 16 synthetic tests at dimensions 10, 20, 50, and 100, GP-LCB with a 400-point expert cap achieved a mean log-scale simple-regret score of 3.975 versus TuRBO’s 4.080. Results cover benchmarks and simulations, not real-world experiments.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.