[15a-A205-1~10] 23.1 Joint Session N "Informatics"

[15a-A205-9] Prediction of electron density from atomic configuration using machine learning

〇Kiyou Shibata^1,2, Teruyasu Mizoguchi^1,2 (1.IIS, the Univ. of Tokyo, 2.Grad. Sch. of Eng., the Univ. of Tokyo)

Keywords:machine learning, electron density of state, materials informatics

Electron density of state (DOS) is the basis for consideration of various physical properties and chemical reactions. DOSs are usually calculated by first-principles calculations, which require a severe computational workload. Substituting first-principles calculations with a faster DOS prediction method will accelerate searching new materials and understanding physical properties. Here we propose prediction of projected DOS on metallic atom nanoclusters from atomic configurations using machine learning.