BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IMX Talks - More than physics\, more than data: integrated machine
 -learning models for materials
DTSTART:20240625T130000
DTEND:20240625T140000
DTSTAMP:20260415T111815Z
UID:3bd58adfdea26390c44de2f598fd935b9da8658bbfc9c9baead1e661
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Michele Ceriotti\, COSMO\nMachine-learning techniques ar
 e often applied to perform "end-to-end" predictions\, that is to make a bl
 ack-box estimate of a property of interest using only a coarse description
  of the corresponding inputs.\nIn contrast\, atomic-scale modeling of matt
 er is most useful when it allows one to gather a mechanistic insight into 
 the microscopic processes that underlie the behavior of molecules and mate
 rials. \nIn this talk I will provide an overview of the progress that has
  been made combining these two philosophies\, using data-driven techniques
  to build surrogate models of the quantum mechanical behavior of atoms\, e
 nabling "bottom-up" simulations that reveal the behavior of matter in real
 istic conditions with uncompromising accuracy. \nI will discuss two ways 
 by which physical-chemical ideas can be integrated into a machine-learning
  framework. \nOne way involves using physical priors\, such as smoothness
  or symmetry of the structure-property relations\, to inform the mathemati
 cal structure of a generic ML approximation. The other entails a deeper le
 vel of integration\, in which explicit physics-based models and approximat
 ions are built into the model architecture. \nI will discuss several exam
 ples of the application of these ideas\, from the calculation of electroni
 c excitations to the design of solid-state electrolyte materials for batte
 ries and high-entropy alloys for catalysis\, emphasizing both the accuracy
  and the interpretability that can be achieved with a hybrid modeling appr
 oach\, and providing an overview of the exciting research directions that 
 are made available by these new modeling tools. 
LOCATION:MXF 1 https://plan.epfl.ch/?room==MXF%201 https://epfl.zoom.us/j/
 68483786187
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR