HIGHLIGHTS
- What: The authors assess the state-of-the-art of machine learning in property prediction focusing on three core aspects: data representation and model. More specifically, the authors focus on methods that allow training on multiple datasets. This approach shows satisfactory results, but its performance on certain molecule classes such as radicals remains unclear. This approach has been used to calculate activation energies, as well as pre-exponential factors of reactions (Paraskevas et_al 2015, 2016; Sabbe et_al 2008b, 2010; Van de Vijver et_al 2018).
- Who: Kevin M. Van Geem from the Ghent University, Technologiepark, Gent . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.