Data-driven multiscale methods and their applications in porous media

Publisher:闻天明Release Time:2020-12-08Number of visits:12

Speaker:   Prof. Lijian Jiang 
Time:       10:00-11:00, Dec. 10

Location:  SIST 1A 106

Host:        Qifeng Liao 
  
Abstract:

In the talk, I will present two data-driven multiscale finite element methods: (1) dynamic mode decomposition and deep learning based CEM-GMsFEM for nonlinear multiscale dynamic system; (2) ensemble based GMsFEM for multiscale stochastic problems. Some applications in porous media flows will be presented.

 

Bio: 

 

SIST Seminar 202008