Recurrent Networks and NARMA Modeling

Part of Advances in Neural Information Processing Systems 4 (NIPS 1991)

Bibtex Metadata Paper

Authors

Jerome Connor, Les Atlas, Douglas Martin

Abstract

There exist large classes of time series, such as those with nonlinear moving average components, that are not well modeled by feedforward networks or linear models, but can be modeled by recurrent networks. We show that recurrent neural networks are a type of nonlinear autoregressive-moving average (N ARMA) model. Practical ability will be shown in the results of a competition sponsored by the Puget Sound Power and Light Company, where the recurrent networks gave the best performance on electric load forecasting.