Reproduce “Martingale Posterior Cox Processes” 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
Joshua Rooijakkers, Edwin Fong
MPCP estimates Cox-process intensity by separating expected event count from the density of event locations, then imputing plausible future events to form uncertainty samples without MCMC or variational inference. For Japan-area earthquakes, it reported a held-out joint test log-likelihood of 1094 ± 64.3 and runtime of 48.6 ± 8.76 seconds, outperforming tested variational baselines under memory-limited settings.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.