Reproduce “When Context Compensates for Sparse Event History: AlphaEarth for Spatio-Temporal Point-Process Forecasting” 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
Yahya Aalaila, Mouad Elhamdi, Gerrit Großmann, Daniel Jenson, Elizaveta Semenova, Sebastian Vollmer
The authors test whether AlphaEarth embeddings improve EMS forecasting in held-out regions. They add a 64-dimensional satellite-derived place representation to a matched log-Gaussian Cox process model and compare predictive densities for calls across eight masks in Pennsylvania. Mean density ratios are 5.42× at one week and 2.37× at two weeks, remaining above parity at longer histories; operational gains and generalization beyond this setup are untested.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.