A Parallel Artificial Neural Network Implementation
Ian Wesley-Smith, Baton Rouge, Gabrielle Allen · 2006
Traditional computational methods are highly structured and linear, properties which they derive from the digital nature of computers. These methods are highly effective at solving certain classes of problems: physics simulations, mathematical models, or the analysis of proteins. Classical computational methods are not effective at solving other problems, such as pattern recognition, adaptive learning, and spam filtering. Some biological systems, however, excel at the latter class of problems. For example, the human mind can quickly identify a face, even if it has changed heavily from the last time it was seen, while traditional computational systems are unable to accomplish facial recognition efficiently and accurately even if minor facial or environmental alterations occur. Attempts to create facsimiles of these biological systems electronically have resulted in the creation of artificial neural networks. Similar to their biological counterparts, artificial neural networks are massively parallel systems capable of learning and making generalizations. The inherent parallelism in the network allows for a distributed software implementation of the artificial neural network, causing the network to learn and operate in parallel, theoretically resulting in a performance improvement. This paper will address a parallel neural network implementation in Cactus, a high performance computing framework, the network's relative strengths and weaknesses, and conclude by considering future improvements to the system. 1.