Abstract
The intent of this project is to explore whether the Two Dimensional Chunk Extraction algorithm for the Self-Splitting Modular Neural Network is more effective than the original algorithm. This experiment uses the same data and compares the results of the Two Dimensional Chunk Extraction algorithm versus the original partitioning algorithm, which splits only on one dimension. The result is that the Two Dimensional Chunk Extraction algorithm performs less effectively than the original algorithm on a suite of test problems.