Machine Learning, Phase Stability, and Disorder with the Automatic Flow Framework for Materials Discovery

摘要: Traditional materials discovery approaches - relying primarily on laborious experiments have controlled the pace of technology. Instead, computational offer an accelerated path: high-throughput exploration and characterization virtual structures. These ventures, performed by automated ab-initio frameworks, rapidly expanded volume programmatically-accessible data, cultivating opportunities for data-driven approaches. Herein, a collection robust methods are presented, implemented within Automatic Flow Framework Materials Discovery (AFLOW), that leverages data prediction phase diagrams properties disordered materials. directly address issue synthesizability, bridging gap between simulation experiment. Powering these predictions is AFLOW.org repository inorganic crystals, largest most comprehensive database its kind, containing more than 2 million compounds with about 100 different computed each. As calculated standardized parameter sets, wealth also presents favorable learning environment. Machine algorithms employed property prediction, descriptor development, design rule discovery, identification candidate functional When combined physical models intelligently formulated descriptors, becomes powerful tool, facilitating new applications ranging from high-temperature superconductors to thermoelectrics. been validated synthesis two permanent magnets introduced herein first discovered