A neural network graph partitioning procedure for grid-based domain decomposition

This paper describes a neural network graph partitioning algorithm which partitions unstructured nite ele- ment/volume meshes as a precursor to a parallel domain decomposition solution method. The algorithm works by rst constructing a coarse graph approximation using an automatic graph coarsening method. The coarse graph is partitioned and the results are interpolated onto the original graph to initialize an optimization of the graph partition problem. In practice, a hierarchy of (usually more than two) graphs are used to help obtain the nal graph partition. A mean eld theorem neural network is used to perform all partition optimization. The partitioning method is applied to graphs derived from unstructured nite element meshes and in this context it can be viewed as a multi-grid partitioning method. Copyright