报告题目：Adaptive multi-fidelity surrogatemodeling for Bayesian inference in inverse problems

报 告 人：周涛 中国科学院数学与系统科学研究院

报告时间：2020年6月21日上午 10:00-11:00

报告地点：腾讯会议 ID：503 563 582

会议密码：200621

或添加至会议列表：

http://meeting.tencent.com/s/D8SrYM3jWlOx

校内联系人：张凯 zhangkaimath@mdjtykj.cn

报告摘要：

Thegeneralized polynomial chaos (gPC) are widely used as surrogate models inBayesian inference to speed up the Markov chain Monte Carlo simulations.However, the use of gPC-surrogates introduces model errors that may severelydistort the estimate of the posterior distribution. In this talk, we present anadaptive procedure to construct an adaptive gPC-surrogate. The key idea is torefine the surrogate over a sequence of samples adaptively so that the surrogateis much more accurate in the posterior region. We then introduce an adaptivesurrogate modeling approach based on deep neural networks to handle problemswith high dimensional parameters.

报告人介绍：

TaoZhou is currently an Associate Professor in Chinese Academy of Sciences. Beforejoining CAS, he was a postdoc fellow in EPFL in Switzerland during 2011-2012.Dr. Zhou’s research interests include Uncertainty Quantification (UQ),Parallel-in-Time Algorithms, Spectral Methods and Stochastic Optimal Control.He has published more than 50 papers in top international journals such as SIAMReview, SINUM and JCP. He was a recipient of the NSFC Career Award forExcellent Young Scholars (2018) and CSIAM Excellent Young Scholar Prize (2016).Dr. Zhou serves as Associate Editor for many international journals such asSIAM Journal on Scientific Computing (SISC) and Communications in ComputationalPhysics (CiCP). He also serves as the Associate Editor-in-Chief ofInternational Journal for UQ. Since 2018, he has been the Chief Scientist of ScienceChallenge Project on UQ supported by State Administration of Science,Technology and Industry for National Defense.