报告题目：Big-Data Analytics for Materials Science: Concepts, Challenges, and Hype
报 告 人：Matthias Scheffler（Fritz-Haber-Institut der Max-Planck-Gesellschaft Faradayweg 4-6, D-14195 Berlin, Germany）
报告时间：10月22日，13：30
报告地点：宝山校区E408会议室
邀 请 人：张文清 教授
Abstract： Materials science and engineering is the exploration of how materials behave and how they may be utilized in technological systems. New materials influence all aspects of our society, as they are important in the development of essentially every new commercial product, be it for better or novel solar panels, harder surfaces, better catalysts, and countless other applications.
The number of different materials is very large - virtually infinite. So far we only know very few of those materials, and the potential value of new materials is enormous. On the steady search for advanced or even novel materials with tailored properties and functions, high-throughput screening is by now an established branch of materials research. For successfully exploring the huge chemical-compound space from a computational point of view, two aspects are crucial. These are (i) reliable methodologies to accurately describe all relevant properties for all materials on the same footing, and (ii) new concepts for extracting maximal information from the big data of materials that are produced since many years with an exponential growth rate.
The talk will address both challenges. In particular, I will present a Test Set for Materials Science and Engineering that enables quality control of first-principles calculations. Furthermore, I will demonstrate the possibilities offered by statistical learning theory for big data of materials. Indeed, machine-learning methods can identify structure in big data that is invisible to humans. The talk also emphasizes the importance of causality in the learning process.