Highly versatile and accurate machine learning methods for predicting perovskite properties
Ziming Chen, Jing Wang, Can‐Jie Li, Baiquan Liu, Dongxiang Luo, Yonggang Min, Nianqing Fu, Qifan Xue
Abstract
A dataset of 3720 ABX3-type perovskites and 2660 A 2 B(I)B(II)X 6 -type double perovskites was collected and cleaned up to train a machine learning model that predicts features such as band gaps. SHAP interpretability analysis provides new insights for bandgap evaluation.
Topics & Concepts
Perovskite (structure)Band gapPhotovoltaic systemComputer scienceMaterials scienceArtificial intelligenceMachine learningOptoelectronicsEngineeringElectrical engineeringChemical engineeringPerovskite Materials and Applications