Visit ShanghaiTech University | 中文 | How to find us
HOME > News and Events > Events
Adaptive Gaussian Process Approximation for Bayesian Inference with Expensive Likelihood Functions
Date: 2018/7/26             Browse: 109

Speaker:     Dr. Jinglai Li. SJTU

Time:          14:00—15:00, July 26 

Location:    Room 1A-402, SIST Building

Host:          Prof. Qifeng Liao


We consider Bayesian inference problems with computationally intensive likelihood functions. We propose a Gaussian process (GP) based method to approximate the joint distribution of the unknown parameters and the data. In particular, we write the joint density approximately as a product of an approximate posterior density and an exponentiated GP surrogate. We then provide an adaptive algorithm to construct such an approximation, where an active learning method is used to choose the design points. With numerical examples, we illustrate that the proposed method has competitive performance against existing approaches for Bayesian computation.


Dr. Jinglai Li Received his Bachelor’s degree in Applied Mathematics from the Sun Yat-Sen University in 2002 and his Ph.D degree in mathematics from the State University of New York at Buffalo in 2007. In 2007-2010, he did postdoctoral research in the department of Engineering Science and Applied Mathematics of the Northwestern University. In 2010, he moved to the Massachusetts Institute of Technology, working as a postdoctoral associate at the Department of Aeronautics and Astronautics. He joined the faculty of the Shanghai Jiaotong University in 2012, where he is now a distinguished research fellow at the Institute of Natural Science. He is also holding a visiting faculty position at MIT. Li’s main research interests are scientific computing, computational statistics, uncertainty quantification, as well as the applications in various scientific and engineering problems. His mathematical interests lie on the interface between statistical science and scientific computing. Jinglai Li has published in several scientific journals including the SIAM Journal of Applied Mathematics, Journal of Computational Physics, and Optics Letters. Jinglai Li has also been invited to speak in numerous conference and workshops such as the SIAM Annual Meeting, SIAM Conference of Uncertainty Quantification, SIAM Conference of Computational Science and Engineering.

Research Interests:

Scientific Computing, Computational Statistics, Uncertainty Quantification

SIST-Seminar 18067