Hypothesis

(For Objective 1)

If extended neural learning algorithms are designed for optimized performance with a hardware neural network, it will have similar performance to a fully-connected neural network, and improved performance from a fixed partially-connected neural network. Learning time will be reduced because of the smaller and more efficient synapse layout.












Copyright © Malcolm Stagg 2006. All Rights Reserved.
Website: http://www.virtualsciencefair.org/2006/stag6m2. Email: malcolmst@shaw.ca.