Reproduce “Deep Bayesian Filtering for Nonlinear Structural System Identification: A Tutorial on Structural and Physical Inductive Biases” 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
Wei Liu, Konstantinos Vlachas, Eleni Chatzi
The authors present a tutorial comparing deep Bayesian filtering models for structural monitoring. It explains how learned state-space models infer hidden motion from sensor data, while some incorporate reduced-order physical dynamics. In one held-out simulation, PgDMM’s reconstruction RMSE was 0.0203 versus 0.1020 for its open-loop forecast; model rankings remain uncertain because physics inputs and integration schemes differed.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.