Publications

Discrepancy Modeling Framework: Learning missing physics, modeling systematic residuals, and disambiguating between deterministic and random effects

Published in SIAM Journal on Applied Dynamical Systems, 2022

As data-driven modeling continues to gain momentum, it is imperative that researchers utilize domain knowledge (e.g., first principles physics) to model complex systems. However, in all disciplines, model- measurement mismatch exists. The lack of investigation into this error leaves the missed opportunity to resolve model-measurement mismatch, disambiguate deterministic effects, and improve the underlying model.

Download here