Title:Applications of tensor networks in machine learning and simulation of quantum circuits
Keynote Speaker:Zhang Pan
Tensor networks are commonly used in modeling variational wave functions in quantum many-body physics. In many applications out of physics, the task is to model a joint distribution of many variables, which can be also regarded as a tensor. In this talk I will introduce new approaches on how to use tensor network (including the matrix product states, tree tensor networks, and the PEPS) in modeling probability tensor in graphical models, which finds applications in generative modeling in machine learning, as well as simulations of single-amplitude quantum circuits when converted to a graphical model.
Brief Introduction to the Keynote Speaker:
Zhang Pan, who graduated with a bachelor's degree in 2004 and then was awarded Ph.D. in 2009 by Lanzhou University, has conducted his postdoctoral research in Italy, France, and Santa Fe Institute in the United States. In 2015, he was selected into the "100 Talents Program" of the Chinese Academy of Sciences, and started to serve as an associate researcher and then a researcher at the Institute of Theoretical Physics, Chinese Academy of Sciences, after returning home the same year. His major research orientations are in the interdisciplinary fields of statistical physics and machine learning. For recent years, his research interest has been concentrated in the theory of statistical physics in statistical inference and the new non-supervised machine learning method based on quantum and statistical physics, with multiple papers published in internationally leading journals such as PNAS, PRL, PRX, etc.
Wang Guanghui, Professor at School of Mathematics
9:00, November 12 (Tuesday)
Lecture Hall 1032, Block B, Zhixin Building, Central Campus
Sponsored by: School of Mathematics, Shandong University