1: \begin{abstract}

2: Additive Schwarz algorithms are proposed as an iterative procedure for

3: neural network approximate solutions of partial differential equations.

4: Based on the convergence analysis of the additive Schwarz algorithms

5: in a general Hilbert space setting,

6: the convergence of the neural network approximate solutions

7: is analyzed for the one-level and two-level iterative schemes.

8: Numerical results of the proposed methods are presented for test examples.

9: \end{abstract}