Séminaire
Mercredi 4 Novembre 2026 à 10h00.
Toward the exact exchange-correlation functional: A neural network construct
Prof. GuanHua CHEN
(University of Hong Kong)
Salle de séminaires Lippmann
Invité(e) par
Thomas NIEHAUS
présentera en 2 heures :
''
Deep neural network can be employed to construct exchange−correlation functional. It requires merely the electron densities of small molecules and ions and yet can determine the exchange−correlation functional of large molecules. Comparisons to B3LYP and PBE are given to illustrate the accuracy and transferability of the neural network based XC functional.
''