Neural operator: A new paradigm for learning PDEs by Animashree Anandkumar
Animashree Anandkumar (Caltech/NVIDIA), "Neural operator: A new paradigm for learning PDEs"
http://tensorlab.cms.caltech.edu/users/anima/
Abstract: Partial Differential Equations (PDE) lay the foundation for modeling a wide variety of scientific phenomena. Traditional solvers tend to be slow when high-fidelity solutions are needed. We introduce neural-operator, a data-driven approach that aims to directly learn the solution operator of PDEs. Unlike neural networks that learn function mapping between finite-dimensional spaces, neural operator extends that to learning the operator between infinite-dimensional spaces. This makes the neural operator independent of resolution and grid of training data and allows for zero-shot generalization to higher resolution evaluations. We find that the neural operator is able to solve the Navier-Stokes equation in the turbulent regime with a 1000x speedup compared to traditional methods.
AAAI 2021 Spring Symposium on Combining Artificial Intelligence and Machine Learning with Physics Sciences, March 22-24, 2021 (https://sites.google.com/view/aaai-mlps)
Papers: https://sites.google.com/view/aaai-mlps/proceedings
Slides: https://sites.google.com/view/aaai-mlps/program
Видео Neural operator: A new paradigm for learning PDEs by Animashree Anandkumar канала MLPS - Combining AI and ML with Physics Sciences
http://tensorlab.cms.caltech.edu/users/anima/
Abstract: Partial Differential Equations (PDE) lay the foundation for modeling a wide variety of scientific phenomena. Traditional solvers tend to be slow when high-fidelity solutions are needed. We introduce neural-operator, a data-driven approach that aims to directly learn the solution operator of PDEs. Unlike neural networks that learn function mapping between finite-dimensional spaces, neural operator extends that to learning the operator between infinite-dimensional spaces. This makes the neural operator independent of resolution and grid of training data and allows for zero-shot generalization to higher resolution evaluations. We find that the neural operator is able to solve the Navier-Stokes equation in the turbulent regime with a 1000x speedup compared to traditional methods.
AAAI 2021 Spring Symposium on Combining Artificial Intelligence and Machine Learning with Physics Sciences, March 22-24, 2021 (https://sites.google.com/view/aaai-mlps)
Papers: https://sites.google.com/view/aaai-mlps/proceedings
Slides: https://sites.google.com/view/aaai-mlps/program
Видео Neural operator: A new paradigm for learning PDEs by Animashree Anandkumar канала MLPS - Combining AI and ML with Physics Sciences
Показать
Комментарии отсутствуют
Информация о видео
13 апреля 2021 г. 6:52:56
01:01:10
Другие видео канала
Neural Networks for Solving PDEsSolving PDEs using Machine Learning by Balaji Srinivasan, IIT MadrasAccelerating Simulation of Stiff Nonlinear Systems using Continuous-Time Echo State Nets, RackauckasFourier Neural Operator for Parametric Partial Differential Equations (Paper Explained)An extension of Physics Informed Neural Networks with a modal approach – IVADO Octobre NumériqueZongyi Li's talk on solving PDEs from dataSolving fluid dynamics with neural networks | Vignesh GopakumarAmazon Empire: Anima Anandkumar Interview | FRONTLINEPhysics-informed Neural Networks - Towards solving Navier-Stokes equations (r200526)Machine Learning-3: Linear Algebra (part-1)Application 4 - Solution of PDE/ODE using Neural NetworksPDE 5 | Method of characteristicsIntroduction to Nondimensionalizationمساء dmc - حصرياً ... لقاء رائع ومميز مع الروبوت صوفيا مع الاعلامي اسامة كمال ( اللقاء كامل )Physics Informed Neural Networks : Applying AI in CFDMeshFree Basics Webinar Recordingتعلم الرياضيات في مجال تعلم الآلة و الذكاء الصناعي (أمل سلامي)Finite Element: Nonlinear ProceduresMCQ on Numerical Analysis-01chapter:3/Lec:- 22 Communication system theory class# BIGDATA PHYSICS CLASSES for IIT-JEE /NEET