AI RESEARCH
EqOD: Symmetry-Informed Stability Selection for PDE Identification
arXiv CS.LG
•
ArXi:2605.11524v1 Announce Type: new Data-driven identification of partial differential equations (PDEs) relies on sparse regression over a candidate library of differential operators, where larger libraries inflate false positives under observation noise and smaller libraries risk missing true terms. We