%0 Generic %1 rackauckas2021universal %A Rackauckas, Christopher %A Ma, Yingbo %A Martensen, Julius %A Warner, Collin %A Zubov, Kirill %A Supekar, Rohit %A Skinner, Dominic %A Ramadhan, Ali %A Edelman, Alan %D 2021 %K julialang sciml %R 10.48550/arXiv.2001.04385 %T Universal Differential Equations for Scientific Machine Learning %U https://arxiv.org/abs/2001.04385 %X In the context of science, the well-known adage ä picture is worth a thousand words" might well be ä 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.

@misc{rackauckas2021universal, 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.}, added-at = {2024-02-12T01:54:26.000+0100}, archiveprefix = {arXiv}, author = {Rackauckas, Christopher and Ma, Yingbo and Martensen, Julius and Warner, Collin and Zubov, Kirill and Supekar, Rohit and Skinner, Dominic and Ramadhan, Ali and Edelman, Alan}, biburl = {https://www.bibsonomy.org/bibtex/2e4e328cbcb1bf7f9285b18deaa616c4c/gdmcbain}, doi = {10.48550/arXiv.2001.04385}, eprint = {2001.04385}, howpublished = {arXiv:2001.04385 [cs.LG]}, interhash = {4d307969bb133c999196a4b685c9d274}, intrahash = {e4e328cbcb1bf7f9285b18deaa616c4c}, keywords = {julialang sciml}, primaryclass = {cs.LG}, timestamp = {2024-02-12T01:54:26.000+0100}, title = {Universal Differential Equations for Scientific Machine Learning}, url = {https://arxiv.org/abs/2001.04385}, year = 2021 }