A Review on Background and Applications of Machine Learning in Materials Research

Abstract

In recent decades, Artificial Intelligence (AI) has garnered considerable interest owing to its potential to facilitate greater levels of automation and speed up overall output. There has been a significant increase in the quantity of training data sets, processing capacity, and deep learning techniques that are all favorable to the widespread use of AI in fields like material science. Attempting to learn anything new by trial and error is a slow and ineffective approach. Therefore, AI, and particularly machine learning, may hasten the process by gleaning rules from information and constructing predictive models. In traditional computational chemistry, human scientists give the formulae, and the computer just crunches the numbers. In this article, we take a look back at the ways in which artificial intelligence has been put to use in the creation of new materials, such as in their design, performance prediction, and synthesis. In these programs, an emphasis is placed on the specifics of AI methodology implementation and the benefits gained over more traditional approaches. The last section elaborates, from both an algorithmic and an infrastructural perspective, where AI is headed in the future.