Components for a sequence processing neural network
David A. Calvert, Deborah Stacey · 2003
Describes the extension of an artificial neural network for sequence modeling and recall. This basic system learns an initial data set which it uses as an exemplar for later sequence processing. This system is designed to take full advantage of the features common to artificial neural networks. The extension to this system removes a limitation found in the initial architecture which prevented it from predicting sequence values from outside the range of the training dataset.