Abstract
Self-splitting neural networks provide a new method for solving complex problems by using multiple neural networks in a divide-and-conquer approach to reduce the domain space each network must solve. However, choosing optimal points for splitting the domain is a difficult problem. A visualization tool exists to help understand how splitting occurs in the self-splitting neural network. This project provided several new enhancements to the tool to expand its scope and improve existing functionality. These enhancements included a new extensible framework for adding additional learning methods to the algorithm, integrating enhancements to the algorithm that had been discovered since the original tool was released, and several new features for observing how the domain space is partitioned. These modifications can be used to develop further insights into the splitting and training processes.