Advances of machine learning in materials science: Ideas and techniques

Abstract

In this big data era, the use of large dataset in conjunction with machine learning (ML) has been increasingly popular in both industry and academia. In recent times, the field of materials science is also undergoing a big data revolution, with large database and repositories appearing everywhere. Traditionally, materials science is a trial-and-error field, in both the computational and experimental departments. With the advent of machine learning-based techniques, there has been a paradigm shift: materials can now be screened quickly using ML models and even generated based on materials with similar properties; ML has also quietly infiltrated many sub-disciplinary under materials science. However, ML remains relatively new to the field and is expanding its wing quickly. There are a plethora of readily-available big data architectures and abundance of ML models and software; The call to integrate all these elements in a comprehensive research procedure is becoming an important direction of material science research. In this review, we attempt to provide an introduction and reference of ML to materials scientists, covering as much as possible the commonly used methods and applications, and discussing the future possibilities.

Universal scales of electronegativity and ionicity from electron scattering factors

“…We hope that the newly proposed scattering electronegativity and ionicity scales, together with the proposed critical ionicity gap and the Mg rule, will trigger more thinking and exploration in the future, and may facilitate the development of new statistical methods for materials or the discovery of new trends in physical chemistry. We also expect that our proposed electronegativity and ionicity scales may be useful in machine learning 45,46 or data mining in chemistry, physics, energy, and materials science.…”

Local symmetry-driven interfacial magnetization and electronic states in (ZnO)n/(w-FeO)n superlattices

“…The interesting electronic and magnetic features of the (ZnO) n /(w-FeO) n superlattices may shed some light on the design of novel devices with a combination of semiconductors and functional materials such as multiferroics. Furthermore, given the rapid development of machine learning in materials science, 61 one can expect that research on superlattices including (ZnO) n /(w-FeO) n and other semiconductor/multiferroic superlattices or interfaces can benefit from advances in artificial intelligence.…”

Interfacial electronic state between hexagonal ZnO and cubic NiO

“…33 If such interesting local-symmetry-induced interfaces can be realized experimentally, the model proposed in this work will be well-suited for examining interface properties. With the advancement of characterization techniques, 25,34–37 and the development of materials informatics and machine learning, 38 the combination of our proposed model and big data in materials science may contribute to the in-depth study of interfaces.…”