Machine learning approach to explore higher Young modulus materials with first principle calculation

Recently, data-driven approaches to design materials, so called ``materials informatics'', are attracting more attentions. Particularly, numerical approaches combining the first principle calculation and machine learning methods are intensively advanced. We apply the method combining the Bayesian optimization method and the first principle calculation for exploring the most rigid materials in certain binary hexagonal compounds. As the result, the combining method succeed in finding the best materials in the search space and the dataset including various type data makes the exploring process efficient.