A service provided by the WU Library and the WU IT-Services

Stationary and integrated autoregressive neural network processes

Trapletti, Adrian and Leisch, Friedrich and Hornik, Kurt (1998) Stationary and integrated autoregressive neural network processes. Working Papers SFB "Adaptive Information Systems and Modelling in Economics and Management Science", 24. SFB Adaptive Information Systems and Modelling in Economics and Management Science, WU Vienna University of Economics and Business, Vienna.

[img]
Preview
PDF
Download (335Kb) | Preview

Abstract

We consider autoregressive neural network (ARNN) processes driven by additive noise. Sufficient conditions on the network weights (parameters) are derived for the ergodicity and stationarity of the process. It is shown that essentially the linear part of the ARNN process determines whether the overall process is stationary. A generalization to the case of integrated ARNN processes is given. Least squares training (estimation) of the stationary models and testing for non-stationarity are discussed. The estimators are shown to be consistent and expressions on the limiting distributions are given.

Item Type: Paper
Keywords: neuronales Netz / autoregressiver Prozess / Zeitreihenanalyse / Prognose
Divisions: Departments > Finance, Accounting and Statistics > Statistics and Mathematics
Departments > Informationsverarbeitung u Prozessmanag. > Informationswirtschaft
Departments > Informationsverarbeitung u Prozessmanag. > Produktionsmanagement > Taudes
Departments > Marketing > Service Marketing und Tourismus
Depositing User: Repository Administrator
Date Deposited: 22 Mar 2002 12:51
Last Modified: 27 Feb 2017 11:24
URI: http://epub.wu.ac.at/id/eprint/302

Actions

View Item