10 papers · 1 filter
Deep Uzawa for Kinetic Transport with Lagrange-Enforced Boundaries
Charalambos Makridakis, Aaron Pim, Tristan Pryer +1
We propose a neural network framework for solving stationary linear transport equations with inflow boundary conditions. The method represents the solution using a neural network a…
PINN-DG: Residual neural network methods trained with Finite Elements
Georgios Grekas, Charalambos G. Makridakis, Tristan Pryer
Over the past few years, neural network methods have evolved in various directions for approximating partial differential equations (PDEs). A promising new development is the integ…
On the Stability and Convergence of Physics Informed Neural Networks
Dimitrios Gazoulis, Ioannis Gkanis, Charalambos G. Makridakis
Physics Informed Neural Networks is a numerical method which uses neural networks to approximate solutions of partial differential equations. It has received a lot of attention and…
Deep Ritz-Finite Element methods: Neural Network Methods trained with Finite Elements
Georgios Grekas, Charalambos G. Makridakis
While much attention of neural network methods is devoted to high-dimensional PDE problems, in this work we consider methods designed to work for elliptic problems on domains $Ω\s…
A class of Discontinuous Galerkin methods for nonlinear variational problems
Georgios Grekas, Konstantinos Koumatos, Charalambos Makridakis +1
In the context of Discontinuous Galerkin methods, we study approximations of nonlinear variational problems associated with convex energies. We propose element-wise nonconforming f…
Convergence of Discontinuous Galerkin Methods for Quasiconvex and Relaxed Variational Problems
Georgios Grekas, Konstantinos Koumatos, Charalambos Makridakis +1
In this work, we establish that discontinuous Galerkin methods are capable of producing reliable approximations for a broad class of nonlinear variational problems. In particular,…