Stephan Hoyer: "Improving PDE solvers and PDE-constrained optimization with deep learning and di..."
Machine Learning for Physics and the Physics of Learning 2019
Workshop II: Interpretable Learning in Physical Sciences
"Improving PDE solvers and PDE-constrained optimization with deep learning and differentiable programming"
Stephan Hoyer - Google Inc.
Abstract: Deep learning is different from other machine learning methods: it can be blended seamlessly into larger programs that are optimized end-to-end, an emerging paradigm known as “differentiable programming.” This paradigm is particularly advantageous in the physical sciences, because it allows for combining machine learning with the well understood models and numerical methods of scientific computing.
In this talk, I will present two examples [1, 2] of how differentiable programming can enable interpetable learning in the context of simulating partial differential equations. First, I’ll show how deep learning can be used to improve discretizations inside numerical methods for solving partial differential equations, allowing for accurate simulations at much coarser scales. Second, I’ll show how deep learning can be used to reparameterize optimization landscapes in PDE constrained structural optimization problems, significantly increasing the quality of resulting designs.
[1] Bar-Sinai, Y., Hoyer, S., Hickey, J. & Brenner, M. P. Learning data-driven discretizations for partial differential equations. Proceedings of the National Academy of Sciences 201814058 (2019). doi:10.1073/pnas.1814058116
[2] Hoyer, S., Sohl-Dickstein, J. & Greydanus, S. Neural reparameterization improves structural optimization. arXiv [cs.LG] (2019). https://arxiv.org/abs/1909.04240
Institute for Pure and Applied Mathematics, UCLA
October 18, 2019
For more information: http://www.ipam.ucla.edu/mlpws2
Видео Stephan Hoyer: "Improving PDE solvers and PDE-constrained optimization with deep learning and di..." канала Institute for Pure & Applied Mathematics (IPAM)
Workshop II: Interpretable Learning in Physical Sciences
"Improving PDE solvers and PDE-constrained optimization with deep learning and differentiable programming"
Stephan Hoyer - Google Inc.
Abstract: Deep learning is different from other machine learning methods: it can be blended seamlessly into larger programs that are optimized end-to-end, an emerging paradigm known as “differentiable programming.” This paradigm is particularly advantageous in the physical sciences, because it allows for combining machine learning with the well understood models and numerical methods of scientific computing.
In this talk, I will present two examples [1, 2] of how differentiable programming can enable interpetable learning in the context of simulating partial differential equations. First, I’ll show how deep learning can be used to improve discretizations inside numerical methods for solving partial differential equations, allowing for accurate simulations at much coarser scales. Second, I’ll show how deep learning can be used to reparameterize optimization landscapes in PDE constrained structural optimization problems, significantly increasing the quality of resulting designs.
[1] Bar-Sinai, Y., Hoyer, S., Hickey, J. & Brenner, M. P. Learning data-driven discretizations for partial differential equations. Proceedings of the National Academy of Sciences 201814058 (2019). doi:10.1073/pnas.1814058116
[2] Hoyer, S., Sohl-Dickstein, J. & Greydanus, S. Neural reparameterization improves structural optimization. arXiv [cs.LG] (2019). https://arxiv.org/abs/1909.04240
Institute for Pure and Applied Mathematics, UCLA
October 18, 2019
For more information: http://www.ipam.ucla.edu/mlpws2
Видео Stephan Hoyer: "Improving PDE solvers and PDE-constrained optimization with deep learning and di..." канала Institute for Pure & Applied Mathematics (IPAM)
Показать
Комментарии отсутствуют
Информация о видео
6 ноября 2019 г. 0:43:29
00:53:07
Другие видео канала
Iain Murray: "Introduction to MCMC for Deep Learning"DDPS | "When and why physics-informed neural networks fail to train" by Paris PerdikarisPDE-constrained Optimization Using PETSc/TAO ǀ Alp Dener, Argonne National LaboratoryDavid Duvenaud | Reflecting on Neural ODEs | NeurIPS 2019Automatic Differentiation.But what is a partial differential equation? | DE2Feynman's Lost Lecture (ft. 3Blue1Brown)Stéphane Mallat - Learning Physics with Deep Neural Networks (October 17, 2018)Deep Credit Risk Ranking with LSTM (Kyle Grove)Neural Networks for Solving PDEsLars Ruthotto: "Deep Neural Networks Motivated By Differential Equations (Part 1/2)"Petros Koumoutsakos: "Machine Learning for Fluid Mechanics"Quantitative Bicycles. Deep Learning DesignOptimised Finite Difference Computation from Symbolic Equations | SciPy 2017 | Michael LangeMachine learning comes to Partial Differential EquationsHarvard AM205 video 4.12 - PDE-constrained optimizationDeep learning for technical computations and equation solvingPhysics informed neural networks..Introduction to Nonlinear PDEs I. Nonlinear Diffusion Equation