Anima Anandkumar - Neural operator: A new paradigm for learning PDEs
Talk starts at 1:50
Prof. Anima Anandkumar from Caltech/NVIDIA speaking in the Data-Driven Methods for Science and Engineering Seminar on March 5, 2021
For more information including past and upcoming talks, visit: http://www.databookuw.com/seminars/
Sign up for notifications of future talks: https://mailman.u.washington.edu/mailman/listinfo/datadriven-seminar
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 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.
http://tensorlab.cms.caltech.edu/users/anima/
Видео Anima Anandkumar - Neural operator: A new paradigm for learning PDEs канала Physics Informed Machine Learning
Prof. Anima Anandkumar from Caltech/NVIDIA speaking in the Data-Driven Methods for Science and Engineering Seminar on March 5, 2021
For more information including past and upcoming talks, visit: http://www.databookuw.com/seminars/
Sign up for notifications of future talks: https://mailman.u.washington.edu/mailman/listinfo/datadriven-seminar
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 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.
http://tensorlab.cms.caltech.edu/users/anima/
Видео Anima Anandkumar - Neural operator: A new paradigm for learning PDEs канала Physics Informed Machine Learning
Показать
Комментарии отсутствуют
Информация о видео
6 марта 2021 г. 0:46:32
00:59:56
Другие видео канала
Fourier Neural Operator for Parametric Partial Differential Equations (Paper Explained)How Deep Neural Networks WorkNeural operator: A new paradigm for learning PDEs by Animashree AnandkumarMSML2020 Invited Talk by Prof. George Karniadakis, Brown UniversityTrinity of Artificial Intelligence | Anima Anandkumar | TEDxIndianaUniversityGeorge Karniadakis - From PINNs to DeepOnetsFourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains (10min talk)Anima Anandkumar - AI4Science: A Revolution in the Making (April 21, 2021)Neural Network In 5 Minutes | What Is A Neural Network? | How Neural Networks Work | SimplilearnDeepMind’s New AI Finally Enters The Real World! 🤖Zongyi Li's talk on solving PDEs from dataBut what is a neural network? | Chapter 1, Deep learningDeepXDE: A Deep Learning Library for Solving Differential Equations by Lu LuTransformer Neural Networks - EXPLAINED! (Attention is all you need)Machine Learning of Partial Differential EquationsSpiking Neural Networks for More Efficient AI AlgorithmsPhysics-Informed Neural NetworkDeep Learning to Discover Coordinates for Dynamics: Autoencoders & Physics Informed Machine LearningA Hands-on Introduction to Physics-informed Machine LearningDynamics of Deep Neural Networks--A Fourier Analysis Perspective-Yaoyu Zhang