Recurrent Networks and NARMA Modeling
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.
Cite
Text
Connor et al. "Recurrent Networks and NARMA Modeling." Neural Information Processing Systems, 1991.Markdown
[Connor et al. "Recurrent Networks and NARMA Modeling." Neural Information Processing Systems, 1991.](https://mlanthology.org/neurips/1991/connor1991neurips-recurrent/)BibTeX
@inproceedings{connor1991neurips-recurrent,
title = {{Recurrent Networks and NARMA Modeling}},
author = {Connor, Jerome and Atlas, Les E. and Martin, Douglas R.},
booktitle = {Neural Information Processing Systems},
year = {1991},
pages = {301-308},
url = {https://mlanthology.org/neurips/1991/connor1991neurips-recurrent/}
}