Reproduce “Particle-Gibbs Sampling For Bayesian Feature Allocation 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

Alexandre Bouchard-Côté, Andrew Roth

Particle Gibbs jointly updates rows in Bayesian feature-allocation models using particles to build a row without enumerating all 2^K assignments, giving linear complexity in K for fixed particle count. In a simulated modified PyClone experiment (FBB, D=4, K=4, N=200), DPF’s mean B-Cubed F-measure exceeded Gibbs by 0.2967 after 100 seconds. This finding is limited to synthetic tests.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.