Загрузка страницы

Christoph Ortner - Atomic Cluster Expansion with and Without Atoms - IPAM at UCLA

Recorded 19 April 2023. Christoph Ortner of the University of British Columbia presents "Atomic Cluster Expansion with and Without Atoms" at IPAM's workshop for Scale-Bridging Materials Modeling at Extreme Computational Scales.
Abstract: I will outline a geometric learning framework, the atomic cluster expansion (ACE), which focuses on linear and shallow models. The framework is particularly well-suited for studying particle systems where it often achieves state-of-the-art accuracy / computational performance and such it is well suited for applications requiring high performance. I will first review the general setting, i.e., how the ACE concept can be adapted to provide models in many different contexts, including molecular simulation, electronic structure, wave functions, and particle physics.
In the second part of my talk I will then review an emerging framework or workflow - hyper-active learning - how to efficiently generate training data for ACE models in an automated way. Relatively small and linear ACE models are ideally suited to this setting. I will also go into some details of the priors we impose that make the workflow as robust as possible.
Learn more online at: http://www.ipam.ucla.edu/programs/workshops/workshop-ii-scale-bridging-materials-modeling-at-extreme-computational-scales/

Видео Christoph Ortner - Atomic Cluster Expansion with and Without Atoms - IPAM at UCLA канала Institute for Pure & Applied Mathematics (IPAM)
Показать
Комментарии отсутствуют
Введите заголовок:

Введите адрес ссылки:

Введите адрес видео с YouTube:

Зарегистрируйтесь или войдите с
Информация о видео
20 апреля 2023 г. 1:35:49
00:53:04
Другие видео канала
Ryan Babbush - Searching for valuable applications of fault-tolerant quantum computers in chemistryRyan Babbush - Searching for valuable applications of fault-tolerant quantum computers in chemistryJustin Smith - The state of neural network interatomic potentials - IPAM at UCLAJustin Smith - The state of neural network interatomic potentials - IPAM at UCLARichard Hennig & Jason Gibson - AI-driven workflows for the discovery of novel superconductorsRichard Hennig & Jason Gibson - AI-driven workflows for the discovery of novel superconductorsJames Corbett - Flux: a next generation resource manager for HPC and beyond - IPAM at UCLAJames Corbett - Flux: a next generation resource manager for HPC and beyond - IPAM at UCLAJuliane Mueller - Adaptive Computing and multi-fidelity learning - IPAM at UCLAJuliane Mueller - Adaptive Computing and multi-fidelity learning - IPAM at UCLAMichele Ceriotti - Machine learning for atomic-scale modeling - potentials and beyond - IPAM at UCLAMichele Ceriotti - Machine learning for atomic-scale modeling - potentials and beyond - IPAM at UCLAAurora Clark - high-dimension perspective on extracting & encoding information in chemical systemsAurora Clark - high-dimension perspective on extracting & encoding information in chemical systemsRalf Drautz - From electrons to the simulation of materials - IPAM at UCLARalf Drautz - From electrons to the simulation of materials - IPAM at UCLAJoshua Schrier - Creating Complex Scientific Workflows that Reach into the Real World - IPAM at UCLAJoshua Schrier - Creating Complex Scientific Workflows that Reach into the Real World - IPAM at UCLASamuel Blau - High-Throughput DFT and Monte Carlo for Reaction Networks and Machine LearningSamuel Blau - High-Throughput DFT and Monte Carlo for Reaction Networks and Machine LearningAmit Acharya - Slow time-scale behavior of fast microscopic dynamics - IPAM at UCLAAmit Acharya - Slow time-scale behavior of fast microscopic dynamics - IPAM at UCLAFrederic Legoll - Parareal algorithms for molecular dynamics simulations - IPAM at UCLAFrederic Legoll - Parareal algorithms for molecular dynamics simulations - IPAM at UCLABoris Kozinsky - Uncertainty-aware machine learning models of many-body atomic interactionsBoris Kozinsky - Uncertainty-aware machine learning models of many-body atomic interactionsThomas Swinburne - Learning uncertainty-aware models of defect kinetics at scale - IPAM at UCLAThomas Swinburne - Learning uncertainty-aware models of defect kinetics at scale - IPAM at UCLAJames Kermode - Multiscale and data-driven methods for the simulation of material failureJames Kermode - Multiscale and data-driven methods for the simulation of material failureRobert Lipton - Fracture as an emergent phenomenon - IPAM at UCLARobert Lipton - Fracture as an emergent phenomenon - IPAM at UCLAYekaterina Epshteyn - Multiscale modeling and analysis of grain growth in polycrystalline materialsYekaterina Epshteyn - Multiscale modeling and analysis of grain growth in polycrystalline materialsFlorin Bobaru - Peridynamic fracture across scales: large scale computations with fast methodsFlorin Bobaru - Peridynamic fracture across scales: large scale computations with fast methodsXiantao Li - A stochastic algorithm for self-consistent calculations in DFT - IPAM at UCLAXiantao Li - A stochastic algorithm for self-consistent calculations in DFT - IPAM at UCLAGabor Csányi - Machine learning potentials: from polynomials to message passing networksGabor Csányi - Machine learning potentials: from polynomials to message passing networks
Яндекс.Метрика