Machine Learning at the Atomic Scale: From Structural Representations to Chemical Insights

  • TH Department Online Seminar
  • Date: May 6, 2022
  • Time: 14:00
  • Speaker: Professor Michele Ceriotti
  • École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
  • Location: https://zoom.us/j/95871552113?pwd=Ti8rZTdIRzlHSFpWQ2xBZVJOc3lQUT09
  • Room: Meeting ID: 958 7155 2113 | Passcode: 889419
  • Host: TH Department
Machine Learning at the Atomic Scale: From Structural Representations to Chemical Insights
When modeling materials and molecules at the atomic scale, achieving a realistic level of complexity and making quantitative predictions are usually conflicting goals. Data-driven techniques have made great strides towards enabling simulations of materials in realistic conditions with uncompromising accuracy.

In particular, statistical regression techniques have become very fashionable as a tool to predict the properties of systems at the atomic scale, sidestepping much of the computational cost of accurate quantum chemical calculations, and making it possible to perform simulations that require thorough statistical sampling without compromising on the accuracy of the electronic structure model.
I will show that data-driven modelling can be rooted in a mathematically rigorous and physically-motivated symmetry-adapted framework, and discuss the benefits of taking a well-principled approach.
I will present several examples demonstrating how the combination of machine-learning and atomistic simulations can offer useful insights on the behavior of complex systems, and discuss the challenges towards an integrated modeling framework in which physics-driven and data-driven steps can be combined to improve the accuracy, the computational efficiency and the transferability of predictions, from interatomic potentials to electronic-structure properties.

Go to Editor View