Title: A hybrid intelligent system for forecasting crude oil price

Authors: Mohsen Mehrara; Hamid Abrishami; Mehdi Ahrari; Vida Varahrami

Addresses: Faculty of Economics, University of Tehran, Kargar-e-Shomali Avenue, P.O. Box 14166-6445, Tehran 14117, Iran. ' Faculty of Economics, University of Tehran, Kargar-e-Shomali Avenue, P.O. Box 14166-6445, Tehran 14117, Iran. ' Faculty of Economics, University of Tehran, Kargar-e-Shomali Avenue, P.O. Box 14166-6445, Tehran 14117, Iran. ' Faculty of Economics, University of Tehran, Kargar-e-Shomali Avenue, P.O. Box 14166-6445, Tehran 14117, Iran

Abstract: In this paper, a novel hybrid intelligent framework is developed by applying a systematic integration of group method of data handling (GMDH) neural networks with genetic algorithm and rule-based expert system with web-based text mining for crude oil price forecasting. Our research reveals that employing a hybrid intelligent framework for crude oil price forecasting provides more accurate results than those obtained from GMDH neural networks when reviewing empirical data from this recent period of financial crisis and results will be so better when we employ hybrid intelligent system with generalised autoregressive conditional heteroskedasticity (GARCH) for crude oil price volatility forecasting. We can use from this method for other industries (gas, coal, ethanol, etc.).

Keywords: crude oil; price forecasting; WTM; web-based text mining; GMDH; group method of data handling; neural networks; GAs; genetic algorithms; hybrid intelligent systems; RES; rule-based expert systems; GARCH; generalised autoregressive conditional heteroskedasticity; price volatility; intelligent forecasting; oil prices.

DOI: 10.1504/IJEBR.2013.050639

International Journal of Economics and Business Research, 2013 Vol.5 No.1, pp.1 - 16

Received: 08 May 2021
Accepted: 12 May 2021

Published online: 08 Nov 2012 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article