Universal Differential Equations for Scientific Machine Learning
arXiv:2001.04385
Abstract
In the context of science, the well-known adage "a picture is worth a thousand words" might well be "a model is worth a thousand datasets." In this manuscript we introduce the SciML software ecosystem as a tool for mixing the information of physical laws and scientific models with data-driven machine learning approaches. We describe a mathematical object, which we denote universal differential equations (UDEs), as the unifying framework connecting the ecosystem. We show how a wide variety of applications, from automatically discovering biological mechanisms to solving high-dimensional Hamilton-Jacobi-Bellman equations, can be phrased and efficiently handled through the UDE formalism and its tooling. We demonstrate the generality of the software tooling to handle stochasticity, delays, and implicit constraints. This funnels the wide variety of SciML applications into a core set of training mechanisms which are highly optimized, stabilized for stiff equations, and compatible with distributed parallelism and GPU accelerators.
5 figures, 2 tables, 11 supplemental figures, 29 pages, 25 supplemental pages
Cited by in corpus (10)
- Differentiable modeling to unify machine learning and physical models and advance Geosciences
- Multi-resolution partial differential equations preserved learning framework for spatiotemporal dynamics
- Discovering Sparse Interpretable Dynamics from Partial Observations
- Performance of Julia for High Energy Physics Analyses
- Uncertainty quantification in mechanistic epidemic models via cross-entropy approximate Bayesian computation
- Hybrid analysis and modeling, eclecticism, and multifidelity computing toward digital twin revolution
- Data-driven discovery of dynamics from time-resolved coherent scattering
- Neural Infalling Cloud Equations (NICE): Increasing the Efficacy of Subgrid Models and Scientific Equation Discovery using Neural ODEs and Symbolic Regression
- NORi: An ML-Augmented Ocean Boundary Layer Parameterization
- Phase State Space Models: Parallel, Surrogate-Free Training of Spiking Networks