Appendices A The Delta Rule of Backpropagation

This paper presents a derivation of a training algorithm for backpropagation neural networks which operate upon time-dependent input data, called “temporal backpropagat ion.” The derivation uses transformation rules to generate the algorithm directly from the standard backpropagation a lgorithm. Issues encountered when training time-delay neural networks are discussed, and a number of examples are s hown.