Id: arxiv/2607.13574 Url: https://arxiv.org/abs/2607.13574 Tags: math.NA, cs.LG, math.AP Score: 0 Title: Approximation of solutions of parameter-dependent problems by residual neural networks Domain: math.NA Source: arxiv Authors: Ana Carpio Reviews: 0 Subject: math.NA Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-07-15 Page: https://delphi.science/r/arxiv/2607.13574 We develop a convergent scheme to train neural networks involving analytic activation functions based on gradient flows. Convergence properties are guaranteed by Lojasiewicz theory. The main advantage of this approach is its simplicity of implementation. The coefficients of the network are approximated by solving a system of ordinary differential equations. We test the method by constructing residual neural network approximations of solutions of parametric problems. The dependence of the solutions of simple ordinary differential equations on a few parameters is correctly reproduced. The…