Efficient autonomous material search method combining <i>ab</i> initio calculations, autoencoder, and multi-objective Bayesian optimization

Autonomous material search systems that combine ab initio calculations and Bayesian optimization are very promising for exploring huge spaces. Setting up an appropriate space is necessary efficient autonomous search. However, performing the within set using prepared descriptors not sufficient to obtain search, which can be achieved by prioritizing specific properties. Here, a system autonomously while multi-objective considers similarities among elements emphasizes proposed. This has been used exploration of Heusler alloys. The successfully proposed several candidate materials with half-metallic versatile applied various properties systems.