Artificial intelligence-based new material design

Abstract

For a long time, design of new materials was implemented on the basis of costly and time-consuming experimental approach. Nowadays machine learning, data mining and Big Data techniques allow to develop a new approach, a so-called data-driven design of new materials. In this paper, we consider an approach to alloy discovery that is based on a synergy of deep learning and fuzzy logic methods. The approach allows to design materials with predefined characteristics by computer-aided generation of the underlying crystal structures and optimization of their parameters. An example is provided to illustrate the approach.