Current browse context:
cs.LG
Change to browse by:
References & Citations
Mathematics > Numerical Analysis
Title: Deep Ritz revisited
(Submitted on 9 Dec 2019 (v1), last revised 10 Jan 2020 (this version, v2))
Abstract: Recently, progress has been made in the application of neural networks to the numerical analysis of partial differential equations (PDEs). In the latter the variational formulation of the Poisson problem is used in order to obtain an objective function - a regularised Dirichlet energy - that was used for the optimisation of some neural networks. In this notes we use the notion of $\Gamma$-convergence to show that ReLU networks of growing architecture that are trained with respect to suitably regularised Dirichlet energies converge to the true solution of the Poisson problem. We discuss how this approach generalises to arbitrary variational problems under certain universality assumptions of neural networks and see that this covers some nonlinear stationary PDEs like the $p$-Laplace.
Submission history
From: Johannes Müller [view email][v1] Mon, 9 Dec 2019 09:59:38 GMT (16kb)
[v2] Fri, 10 Jan 2020 19:25:16 GMT (16kb)
Link back to: arXiv, form interface, contact.