Abstract

Collective quasiparticle states (molecular orbitals, plasmons, phonons, polarons, excitons, etc.) are often employed to study and understand many particle systems in quantum mechanics (especially in extended systems). Can machine learning (ML) techniques generate such quasiparticle states or approximations thereof, given only the atomistic Hamiltonian as an input and macroscopic observables as an outputya On a larger scale and going toward materials design (materials genomics): How can one generate the necessary and sufficient data to use ML approaches to infer the important collective variables (materials genes, scaling relations, etc.)ya This workshop will gather experts in electronic structure methods together with mathematicians and computer scientists who are interested in or have already been applying ML methods in physics, chemistry, and materials sciences.