A delay damage model selection algorithm for NARX neural networks

Author

Lin, Tsung-Nan ; Giles, C. Lee ; Horne, Bill G. ; Kung, Sun-Yuan

Author_Institution

NEC Res. Inst., Princeton, NJ, USA

Volume

45

Issue

11

fYear

1997

fDate

11/1/1997 12:00:00 AM

Firstpage

2719

Lastpage

2730

Abstract

Recurrent neural networks have become popular models for system identification and time series prediction. Nonlinear autoregressive models with exogenous inputs (NARX) neural network models are a popular subclass of recurrent networks and have been used in many applications. Although embedded memory can be found in all recurrent network models, it is particularly prominent in NARX models. We show that using intelligent memory order selection through pruning and good initial heuristics significantly improves the generalization and predictive performance of these nonlinear systems on problems as diverse as grammatical inference and time series prediction

Keywords

autoregressive processes; delays; identification; inference mechanisms; nonlinear systems; prediction theory; recurrent neural nets; time series; NARX neural networks; delay damage model selection algorithm; embedded memory; generalization; grammatical inference; initial heuristics; intelligent memory order selection; nonlinear autoregressive models with exogenous inputs; nonlinear systems; predictive performance; pruning; recurrent neural networks; system identification; time series prediction; Computer architecture; Computer networks; Memory architecture; National electric code; Neural networks; Neurons; Nonlinear systems; Predictive models; Recurrent neural networks; System identification;

fLanguage

English

Journal_Title

Signal Processing, IEEE Transactions on

Publisher

ieee

ISSN

1053-587X

Type

jour

DOI

10.1109/78.650098

Filename

650098