DYSCO recovers governing equations from noisy high-dim data, accepted at NeurIPS

burny_tech · x · 2026-09-27

Paolo Muratore's DYSCO paper is now accepted at NeurIPS. It tackles recovering the equations governing a dynamical system directly from noisy, high-dimensional observations: it learns latent spaces capturing underlying dynamics and enables recovery of governing equations from them.

Original post →

More from Research

Research channel →