Data Science Methods for Disordered Materials

Disordered materials, such as inorganic glasses, polymer glasses, cement hydrates, amorphous membrane, metal-organic framework glasses, and gels, are critical to our sustainable future. Their design has traditionally been done using the time-consuming and inefficient “trial-and-error” approach. However, new approaches relying on artificial intelligence, machine learning, and topological data analysis offer opportunities for accelerating the discovery of advanced disordered materials. For the planned Ph.D. course, we will introduce data science methods relevant to disordered materials, including various types of machine learning algorithms, data mining and natural language processing methods, persistent homology and so.