Kevin Carlberg - AI for Computational Physics: Toward real-time high-fidelity simulation
Talk starts at 1:30
Dr. Kevin Carlberg speaking in the UW Data-driven methods in science and engineering seminar on April 9, 2021.
For more information including past and upcoming talks, visit: http://www.databookuw.com/seminars/
Sign up for notifications of future talks: https://mailman11.u.washington.edu/mailman/listinfo/datadriven-seminar
Abstract: The explosion of artificial intelligence—especially techniques arising from deep neural networks—has yielded exciting advances in fields such as computer vision, natural language processing, and reinforcement learning. However, the application of these methods to problems in engineering and science remains limited. In this talk, we describe how two particular recent advances in deep learning, namely convolutional autoencoders and long-short-term-memory (LSTM) recurrent neural networks (RNNs) can be employed to overcome two longstanding challenges in nonlinear model reduction: the Kolmogorov limitation of linear subspaces, and accurate error quantification.
Видео Kevin Carlberg - AI for Computational Physics: Toward real-time high-fidelity simulation канала Physics Informed Machine Learning
Dr. Kevin Carlberg speaking in the UW Data-driven methods in science and engineering seminar on April 9, 2021.
For more information including past and upcoming talks, visit: http://www.databookuw.com/seminars/
Sign up for notifications of future talks: https://mailman11.u.washington.edu/mailman/listinfo/datadriven-seminar
Abstract: The explosion of artificial intelligence—especially techniques arising from deep neural networks—has yielded exciting advances in fields such as computer vision, natural language processing, and reinforcement learning. However, the application of these methods to problems in engineering and science remains limited. In this talk, we describe how two particular recent advances in deep learning, namely convolutional autoencoders and long-short-term-memory (LSTM) recurrent neural networks (RNNs) can be employed to overcome two longstanding challenges in nonlinear model reduction: the Kolmogorov limitation of linear subspaces, and accurate error quantification.
Видео Kevin Carlberg - AI for Computational Physics: Toward real-time high-fidelity simulation канала Physics Informed Machine Learning
Показать
Комментарии отсутствуют
Информация о видео
10 апреля 2021 г. 0:34:52
01:10:09
Другие видео канала
George Karniadakis - From PINNs to DeepOnets
Zico Kolter - Incorporating physics and decision making into deep learning via implicit layers
Neural Network 3D Simulation
The first 20 hours -- how to learn anything | Josh Kaufman | TEDxCSU
For the Love of Physics - Walter Lewin - May 16, 2011
David Duvenaud - Handling messy time series with large latent-variable models
Computational Physics with python tutorials- Book Review. Python for physics
Nicholas Zabaras - Physics Informed Learning for Multiscale Dynamical Systems
Anima Anandkumar - Neural operator: A new paradigm for learning PDEs
Wolfram Physics Project Launch
SimuPy: A Python Framework for Modeling and Simulating Dynamical Systems | SciPy 2018 | Margolis
Machine learning may solve computational physics | Max Tegmark and Lex Fridman
Big Data & Computational Physics
But what is a neural network? | Chapter 1, Deep learning
Beyond Deep Learning - Differentiable Programming with Flux - Avik Sengupta | ODSC Europe 2019
Quantum Computing for Computer Scientists
Chris Rackauckas - Physics-Informed Neural Networks (PINNs) | Podcast #42
Beverley McKeon - What’s in a mean (what, how and why)? Towards nonlinear models of wall turbulence
Scientific Machine Learning: Where Physics-based Modeling Meets Data-driven Learning
Machine Learning for Fluid Dynamics: Models and Control