Automated machine learning structure-composition-property relationships of perovskite materials for energy conversion and storage

Perovskite materials are central to the fields of energy conversion and storage, especially for fuel cells. However, they challenged by overcomplexity, coupled with a strong desire new discovery at high speed precision. Herein, we propose approach involving combination extreme feature engineering automated machine learning adaptively learn structure-composition-property relationships perovskite oxide storage. Structure-composition-property between stability other features perovskites investigated. Extreme is used construct great quantity fresh descriptors, crucial subset 23 descriptors acquired sequential forward selection algorithm. The best descriptor determined linear regression. results demonstrate high-efficient non-priori-knowledge investigation materials, providing road discover advanced materials.