Learning Interatomic Potential Energy Surfaces
报告专家：Christoph Ortner 教授（不列颠哥伦比亚大学）
报告时间：4月17日（周五）10:30-11:30
报告地点：四川大学数学学院西202
报告摘要：
The integration of machine learning into the traditional modeling workflows is replacing decades-old ad hoc models with systematic approximations of physical laws and new models that far outstrip their predecessors in accuracy and transferability. A recurring concern in learning interactions in particle systems is the need to incorporate exact preservation of symmetry into the models. This leads to an interesting theoretical question : does symmetry make the approximation problem easier or more difficult, and how can we quantify this? I will give two examples, both motivated by learning surrogate models for molecular dynamics, and study them in an abstract approximation-theoretic setting.
专家简介：
Christoph Orner教授2006年博士毕业于英国牛津大学，之后在牛津大学任教。2014年起任英国华威大学教授。2020年加拿大英属哥伦比亚大学任教授。Ortner教授长期从事材料和分子动力学相关的多尺度方法、深度学习方法等方面的研究，在SIAM Journal on Numerical Analysis, Mathematics of Computation, Archive of Rational Mechanics and Analysis, SIAM Journal on Scientific Computing, Journal of Computational Physics, Multiscale Modeling and Simulation等计算数学、多尺度方法权威期刊发表论文120余篇，曾获得欧盟科学基金科学领军基金、美国-加拿大国家自然科学基金联合重点基金等重要基金项目资助，并担任 Multiscale Modelling and Simulation, IMA Journal of Numerical Analysis, Journal of Computational Mathematics等权威期刊的编委。Ortner教授2015年获得伦敦数学会Whitehead奖，2022年被选为加拿大皇家学会青年会士。
邀请人：王皓
