System identification methods attempt to discover physical models directly from a dataset of measurements, but often there are no guarantees that the resulting models are stable. This video abstract summarizes our recent work that builds in a notion of long-term boundedness (or global stability) for data-driven modeling. Kaptanoglu, Alan A., et al. "Promoting global stability in data-driven models of quadratic nonlinear dynamics." Physical Review Fluids 6.9 (2021): 094401. The arxiv manuscript can also be found here [ Ссылка ].
