Forecasting Economic Data with Neural Networks

Autor:	Farzan Aminian, Daniel T. Walz, M. Aminian, E. Dante Suarez
Rok vydání:	2006
Předmět:	Nonlinear system Economic data Coefficient of determination Artificial neural network Work (electrical) Industrial production Economics Econometrics and Finance (miscellaneous) Linear regression Econometrics Economics Nonlinear regression Computer Science Applications
Zdroj:	Computational Economics. 28:71-88
ISSN:	1572-9974 0927-7099
Popis:	Studies in recent years have attempted to forecast macroeconomic phenomena with neural networks reporting mixed results. This work represents an investigation of this problem using U.S. Real Gross Domestic Production and Industrial Production as case studies. This work is based on a coefficient of determination which accurately measures the ability of linear or nonlinear models to forecast economic data. The significance of our work is twofold: (1) It confirms recent work that neural networks significantly outperform linear regression due to nonlinearities inherent in the data sets, and (2) it provides a systematic approach that guarantees to find the maximum correlation between input(s) and output of interest.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::a94cb1580bc1da7571c5a4fada486602 https://doi.org/10.1007/s10614-006-9041-7 Zobrazit plný text záznamu Full text from SpringerLink