There are two parts to this research. Part 1 is simulation of the neural network on a parallel computer. We discuss some of the design decisions as well as providing code for salient aspects of the model. A pattern recognition example was used as a test case for this simulation. Part 2 is choosing input examplers to achieve network stability. We define various parameters based on the topology of the non-random patterns and based on the experiments performed, suggest ways to choose them to achieve stability of the network.
