Registered Data

[01139] Adaptive sampling and transfer learning techniques for solution of PDEs

  • Session Time & Room : 2E (Aug.22, 17:40-19:20) @G402
  • Type : Contributed Talk
  • Abstract : An adaptive sampling technique applied to the deep Galerkin method (DGM), and separately a transfer learning algorithm also applied to DGM is examined, aimed to improve, and speed up the training of the deep neural network when learning the solution of partial differential equations (PDEs). The proposed algorithms improve the DGM method. The adaptive sampling scheme implementation is natural and efficient. Tests applied to selected PDEs discussing the robustness of our methods are presented.
  • Classification : 35-04, 65-04, Deep learning for the solution of PDEs
  • Format : Online Talk on Zoom
  • Author(s) :
    • Andreas Aristotelous (The University of Akron)