Data-driven Effective Modeling of Multiscale Stochastic Dynamical Systems
This addresses the challenge of modeling complex multiscale systems in fields like physics or engineering, though it appears incremental as it builds on existing data-driven approaches.
The authors tackled the problem of learning the dynamics of slow components in unknown multiscale stochastic systems using bursts of observation data, resulting in a generative stochastic model that accurately captures the effective dynamics in distribution.
We present a numerical method for learning the dynamics of slow components of unknown multiscale stochastic dynamical systems. While the governing equations of the systems are unknown, bursts of observation data of the slow variables are available. By utilizing the observation data, our proposed method is capable of constructing a generative stochastic model that can accurately capture the effective dynamics of the slow variables in distribution. We present a comprehensive set of numerical examples to demonstrate the performance of the proposed method.