Inferring the Structure and Dynamics of Complex Networked Systems

Abstract

Complex systems often contain a large number of nodes that interact with each other. Our ability to observe the behaviors of nodes, often in the form of time-series data, has been advanced spectacularly in recent years, allowing us to extract the dynamic patterns of a system’s individuals. To gain meaningful insights into a real-world system, however, it is necessary to infer its internal network structure and the governing equations of its dynamics from the observation data. To this end, the speaker will talk about the recent progress in data-driven inference of complex system structure and dynamics, especially the deep learning-assisted methods, and discuss the remaining challenges that are worth future effort.

Speaker Intro

严钢，同济大学物理科学与工程学院长聘教授，博士生导师，国家杰出青年基金获得者。2005和2010年于中国科学技术大学分别获得理学学士和工学博士学位。主要研究方向为复杂系统与人工智能交叉领域的理论及应用，研究成果以第一或通讯作者身份发表于Nature, Nature Physics, Nature Computational Science, Physical Review Letters, Physical Review X, National Science Review, AAAI等。近五年主持国家级项目5项，与同行合作获得省部级自然科学一等奖2次。