Rewriting Logic Using Strategies for Neural Networks: An Implementation in Maude

摘要

A general neural network model for rewriting logic is proposed. Thismodel, in the form of a feedforward multilayer net, is represented in rewriting logicalong the lines of several models of parallelism and concurrency that have already beenmapped into it. By combining both a right choice for the representation operations andthe availability of strategies to guide the application of our rules, a new approach forthe classical backpropagation learning algorithm is obtained. An example, the diagno-sis of glaucoma by using campimetric fields and nerve fibres of the retina, is presentedto illustrate the performance and applicability of the proposed model.