1 material found
Author: Audun Skau Hansen
Machine learning in electronic-structure theory
Machine Learning approaches are increasingly being used across many fields of science, with electronic structure theory being no exception. They offer novel approaches to age-old problems. In recent years they have been used in construction of trial wavefunctions, in computing Hamiltonians, and...
Machine learning in electronic-structure theory
https://www.youtube.com/watch?v=Sgp0w74k9kQ
https://pan-training.eu/materials/machine-learning-in-electronic-structure-theory
Machine Learning approaches are increasingly being used across many fields of science, with electronic structure theory being no exception. They offer novel approaches to age-old problems. In recent years they have been used in construction of trial wavefunctions, in computing Hamiltonians, and in direct calculation of properties and forces.
These methods are highly versatile and computationally efficient, yet many questions regarding their interpretability and ability to extrapolate information remains unanswered.
How are they being used in electronic structure theory today, and how do they fit into the bigger picture? Would electronic structure theory have looked anything different if it was conceived in the age of machine learning? This seminar seeks to answer these kinds of questions, and was originally given as a trial lecture at the Hylleraas Centre of Quantum Molecular Sciences at the University of Oslo.
machine learning, electronic-structure theory