Crystal Graph Neural Networks for Data Mining in Materials Science

Technical report, RIMCS LLC 2019 Takenori Yamamoto

Machine learning methods have been employed for materials prediction in various ways. It has recently been proposed that a crystalline material is represented by a multigraph called a crystal graph... (read more)

PDF Abstract

Evaluation Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK COMPARE
Formation Energy OQMD v1.2 CGNN Ensemble MAE 0.0305 # 1
Band Gap OQMD v1.2 CGNN Ensemble MAE 0.0461 # 1
Band Gap OQMD v1.2 CGNN Ensemble AUC 0.9713 # 1
Total Magnetization OQMD v1.2 CGNN Ensemble MAE 0.0691 # 1
Total Magnetization OQMD v1.2 CGNN Ensemble AUC 0.9569 # 1
Materials Screening OQMD v1.2 CGNN Ensemble [email protected] 96 # 1
Materials Screening OQMD v1.2 CGNN Ensemble [email protected] 215 # 1