A Delay Damage Model Selection Algorithm for NARX Neural Networks

Abstract

Recurrent neural networks have become popular models for system identification and time series prediction. NARX (Nonlinear AutoRegressive models with eXogenous inputs) neural network models are a popular subclass of recurrent networks and have beenused in many applications. Though 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.

(Also cross-referenced as UMIACS-TR-96-77)