Samsung Galaxy Z Flip 3 review: The first foldable you may actually want to buy – CNET
November 24, 20219 Best International ETFs To Buy (6 From Vanguard) in 2021
November 24, 2021Is Arctic Zero Healthy? Here’s What You Should Know
November 24, 202111 Foods You Probably Didn’t Know Were Low-Calorie
November 24, 2021There Is Finally a Visible Way Out of the COVID Pandemic
November 24, 2021COVID antiviral pills: what scientists still want to know
November 24, 2021Stanford press release on IGVF
November 24, 2021mentioned in the 2nd to the last paragraph.
https://med.stanford.edu/news/all-news/2021/11/researchers-grants-genome-variations.html
Origins and evolution of extreme life span in Pacific Ocean rockfishes
November 23, 2021Machine-learning-assisted modeling: Physics Today: Vol 74, No 7
November 23, 2021QT : {{”
One of the most successful applications of machine learning to scientific modeling is in MD. Researchers study the properties of materials and molecules by using classical Newtonian dynamics to track the nuclei in a system. One critical issue in MD is how to model the PES that describes the interaction between the nuclei. Traditionally, modelers have dealt with the problem in several ways. One approach, ab initio MD, was developed in 1985 by Roberto Car and Michele
Parrinello7 and computes the interatomic forces on the fly using models based on first principles, such as density functional theory.8 Although the approach accurately describes the system under
consideration, it‘s computationally expensive: The maximum system size that one can handle is limited to thousands of atoms. Another approach uses empirical formulas to model a PES. The method is efficient, but guessing the right formula that can model the PES accurately enough is a difficult task, particularly for complicated systems, such as multicomponent alloys. In 2007 Jörg Behler and Parrinello introduced the idea of using neural networks to model the PES.9 In that new paradigm, a quantum mechanics model generates data that are used to train a neural-network-based PES model.
7. R. Car, M. Parrinello, Phys. Rev. Lett. 55, 2471 (1985).
https://doi.org/10.1103/PhysRevLett.55.2471
9. J. Behler, M. Parrinello, Phys. Rev. Lett. 98, 146401 (2007). https://doi.org/10.1103/PhysRevLett.98.146401
“}}
https://physicstoday.scitation.org/doi/10.1063/PT.3.4793
mentions Parrinelllo classic paper :
#8 has QM & NN (NN models for PES)
linkstream2 microblog 