Tess Smidt: "Euclidean Neural Networks for Emulating Ab Initio Calculations and Generating Atomi..."
Machine Learning for Physics and the Physics of Learning 2019
Workshop I: From Passive to Active: Generative and Reinforcement Learning with Physics
"Euclidean Neural Networks* for Emulating Ab Initio Calculations and Generating Atomic Geometries *also called Tensor Field Networks and 3D Steerable CNNs"
Tess Smidt - Lawrence Berkeley National Laboratory, Physics
Abstract: Atomic systems (molecules, crystals, proteins, nanoclusters, etc.) are naturally represented by a set of coordinates in 3D space labeled by atom type. This is a challenging representation to use for neural networks because the coordinates are sensitive to 3D rotations and translations and there is no canonical orientation or position for these systems. We present a general neural network architecture that naturally handles 3D geometry and operates on the scalar, vector, and tensor fields that characterize physical systems. Our networks are locally equivariant to 3D rotations and translations at every layer. In this talk, we describe how the network achieves these equivariances and demonstrate the capabilities of our network using simple tasks. We’ll also present examples of applying Euclidean networks to applications in quantum chemistry and discuss techniques for using these networks to encode and decode geometry.
Institute for Pure and Applied Mathematics, UCLA
September 27, 2019
For more information: http://www.ipam.ucla.edu/mlpws1
Видео Tess Smidt: "Euclidean Neural Networks for Emulating Ab Initio Calculations and Generating Atomi..." канала Institute for Pure & Applied Mathematics (IPAM)
Workshop I: From Passive to Active: Generative and Reinforcement Learning with Physics
"Euclidean Neural Networks* for Emulating Ab Initio Calculations and Generating Atomic Geometries *also called Tensor Field Networks and 3D Steerable CNNs"
Tess Smidt - Lawrence Berkeley National Laboratory, Physics
Abstract: Atomic systems (molecules, crystals, proteins, nanoclusters, etc.) are naturally represented by a set of coordinates in 3D space labeled by atom type. This is a challenging representation to use for neural networks because the coordinates are sensitive to 3D rotations and translations and there is no canonical orientation or position for these systems. We present a general neural network architecture that naturally handles 3D geometry and operates on the scalar, vector, and tensor fields that characterize physical systems. Our networks are locally equivariant to 3D rotations and translations at every layer. In this talk, we describe how the network achieves these equivariances and demonstrate the capabilities of our network using simple tasks. We’ll also present examples of applying Euclidean networks to applications in quantum chemistry and discuss techniques for using these networks to encode and decode geometry.
Institute for Pure and Applied Mathematics, UCLA
September 27, 2019
For more information: http://www.ipam.ucla.edu/mlpws1
Видео Tess Smidt: "Euclidean Neural Networks for Emulating Ab Initio Calculations and Generating Atomi..." канала Institute for Pure & Applied Mathematics (IPAM)
Показать
Комментарии отсутствуют
Информация о видео
10 октября 2019 г. 2:27:28
00:52:20
Другие видео канала
Risi Kondor: "Fourier space neural networks"What's a Tensor?PSC 2022.02.25 Tess Smidt, MITGauge Equivariant Convolutional Networks and the Icosahedral CNN05 Imperial's Deep learning course: Equivariance and InvarianceIncorporating Symmetry for Learning Spatiotemporal Dynamics - Rose YuStéphane Mallat: "Deep Generative Networks as Inverse Problems"How we teach computers to understand pictures | Fei Fei LiScience vs God: Bryan Enderle at TEDxUCDavisRoto-Translation Covariant Convolutional Networks for (...) - Duits - Workshop 3 - CEB T1 2019e3nn Tutorial MRS 2021 Fall Meeting -- Tutorial 1/6 -- Tess SmidtLeveraging permutation group symmetries for designing equivariant neural networks - Haggai MaronHow Deep Neural Networks WorkXavier Bresson: "Convolutional Neural Networks on Graphs"Week 7 - Symmetry and Equivariance in Neural Networks - Tess SmidtHarmonic Networks: Deep Translation and Rotation EquivarianceBrief Introduction to ab initio Molecular Dynamics (AIMD)Bayesian Deep Learning on a Quantum ComputerMIA: Wouter Boomsma & Jes Frellsen, Convolutional models of molecular structure