Search academic texts

Information × Registration Number 0122U002442, ( 0223U000051  0223U000057  ) R & D request Title New applications of machine learning in predicting unusual material properties Head Bryk Taras M., Доктор фізико-математичних наук Registration Date 09-05-2022 Organization Institute of Condensed Matter Physics of the National Academy of Sciences of Ukraine popup.description1 The project aims at application of a new approach in atomstic computer simulations of disordered materials (metallic mets, glass-forming systems, complex molecular liquids), which are difficult or impossible to describe by simple interatomic interaction potentials. The new approach is based on principle of “machine learning” for estimation of many-parameter correspondence between instantaneous atomic configurations and potential energy surface, which is used then for precise prediction of atoms in the system. By the date the approach of “machine learning” practically is not used in Ukraine. We plan new applications of the “machine learning” in the studies of dynamic phenomena in disordered systems, that practically were not made by the date in this field of science, in particular, for the problems of sound propagation, heat transport, vorticity processes in complex melts, liquids and glasses. popup.nrat_date 2024-12-09 Close