查看原文
其他

【学术视频】机器学习-计算化学Workshop | 麻省理工学院谢天

KouShare 蔻享学术 2021-04-25

更多精彩视频登陆网站

www.koushare.com




图 | 谢天

  目:Tutorials for CGCNN and GDyNets

报告人:谢天

单   位:Massachusetts Institute of Technology

时   间:2019-09-05

地   点:厦门大学化学化工学院

扫码观看精彩报告视频

报告摘要

The exponential increase of open materials data in the past few years poses a challenge to the community to efficiently utilize them for designing better materials. In this talk, I will present our recent works showcasing how graph neural networks can be used as a general framework to utilize large quantities of material data. We will first define the key architecture of crystal graph convolutional neural networks (CGCNN) to represent an arbitrary atomic structure. Then, we will demonstrate how to achieve state-of-the-art performance in predicting material properties and extract material design knowledge from the networks using this framework. Further, we will show several examples of applying this approach to the design of lithium-ion batteries. Finally, we will present our latest work that extends this approach to learn from molecular dynamics simulation trajectories without supervision and understand lithium-ion transport mechanism. 


个人简介

Tian Xie is a PhD student from the department of materials science and engineering at Massachusetts Institute of Technology, advised by Prof. Jeffrey Grossman. He received a B.S. in chemistry from Peking University in 2015. In 2019, he was a PhD software intern at X (formerly Google X), working on an early stage project combing machine learning and physics. After that, he joined DeepMind as a research intern. Tian’s research focuses on the development of machine learning algorithms for materials design, and he works closely with several groups from MIT and CMU to apply those algorithms to the discovery of materials for energy innovations. He developed CGCNN (Crystal Graph Convolutional Neural Networks) as a general approach to represent materials in 2017. 

会议简介

2019年9月3日-6日,由固体表面物理化学国家重点实验室(厦门大学)、福建省理论与计算化学重点实验室和厦门大学化学化工学院主办的“机器学习-计算化学Workshop”在厦门大学化学化工学院举办。本次Workshop邀请了相关领域的研究者报告领域前沿进展,并设置Hands-on tutorials环节帮助学员们熟悉代码的使用。此次Workshop的举办增进了不同领域研究者的交流,促进了开源共享的观念传递,希望推动大数据技术在计算化学和材料模拟等领域的应用。



—— ——往期精彩回顾—— ——

【学术视频】机器学习-计算化学Workshop | 上海大学欧阳润海研究员:Data-Driven Materials Discovery with the Method SISSO

【学术视频】第五届全国光孤子学术研讨会 | 佛蒙特大学杨建科教授:Optical PT Symmetry Theory, Applications and Generalizations

【学术视频】第四届中国能源材料化学研讨会 | 中国科大、中科院大连化物所包信和院士:Journal of Energy Chemistry--with fast rising impact and awareness

【大学频道】上海大学量子与分子结构国际研究中心呈献 | 加州理工学院Nai-Chang Yeh教授Exploring and manipulating different quantum degrees of freedom in two-dimensional van der Waals materials and nano-metamaterials

【大学频道】中国科学院大学卡弗里理论科学研究所呈献 | 西安交通大学刘博教授:Chiral orbital magnetism and topological supersolids in ultracold atoms

【学术视频】ICFSI-17 | 耶拿大学Friedhelm Bechstedt: Validity of Anderson rule for interfaces between 2D semiconductors

【学术视频】中国化学会第十一届全国有机化学学术会议 | 香港中文大学谢作伟院士:A Toolbox for Controlled Functionalization of Carboranes

【学术视频】北京量子磁性暑期学校 | 中国科大邓友金教授:Clock Monte Carlo Methods & Lifted Worm Algorithms

【学术视频】中国化学会第四届中国能源材料化学研讨会 | 南开大学周震教授:钠离子电池和混合电容器的研究进展

● 北大吴飙教授《简明量子力学》PDF电子教材免费下载


END

扫描二维码,关注微信公众号

戳这里,观看精彩视频哟!

    您可能也对以下帖子感兴趣

    文章有问题?点此查看未经处理的缓存