January 25, 2022 10:08
EPFL CIS-RIKEN AIP Joint Seminar #7 20220119 thumbnails


【The 7th Seminar】

Date and Time: January 19th 6:00pm – 7:00pm(JST)
Venue:Zoom webinar

Language: English

Speaker: Michele Ceriotti, EPFL CIS

Title: Representing atoms clouds: the foundations of atomic-scale machine learning

Simulations of matter at the atomic scale are precious to provide a mechanistic understanding of chemical processes, and to design molecules and materials with predictive accuracy. As with many fields of science, machine learning have become an essential part of the modeling toolbox, with many frameworks having become well-established, and many more being developed in new research directions.
Lately, much effort has been dedicated to rationalizing the relationship between models of atomic-scale properties and fundamental physical principles, such as symmetry, locality, and hierarchical decompositions of the interactions between atoms.
I will present an overview of the latest developments, focusing in particular on the problem of representing an atomic structure in a concise, yet complete, fashion. I will discuss a few examples of the implications of these fundamental findings, for both chemical machine learning and in general for problems that require a description of three-dimensional objects in terms of point clouds.