{"id":"arxiv/2603.29863","url":"https://arxiv.org/abs/2603.29863","tags":["math.NA"],"score":0,"title":"Minimum residual discretization of a semilinear elliptic problem","domain":"math.NA","source":"arxiv","authors":["Carlos García Vera","Norbert Heuer","Dirk Praetorius"],"reviews":0,"subject":"math.NA","summary":"We propose a least-squares penalization as a means to extend the discontinuous Petrov-Galerkin (DPG) method with optimal test functions to a class of semilinear elliptic problems. The nonlinear contributions are replaced with independent unknowns so that standard DPG techniques apply to the then linear problem with non-trivial kernel. The nonlinear relations are added as least-squares constraints. Assuming solvability of the semilinear problem and an Aubin-Nitsche-type approximation property for the primal variable, we prove a Cea estimate for the approximation error in canonical norms.…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-31","reactions":{}}