Please use this identifier to cite or link to this item: https://bspace.buid.ac.ae1234/631
Title: Forecasting the Direction of GCC Stock Indexes Using Support Vector Machine (SVM) Algorithm
Authors: Zeino, Yaser Abdullah
Keywords: forecasting
GCC stock indexes
Support Vector Machine (SVM)
stock markets
Artificial Intelligence (AI)
Issue Date: Mar-2014
Publisher: The British University in Dubai (BUiD)
Abstract: Forecasting the direction of stock markets is a very challenging task for investors and decision makers. Recently, investors in stock exchange start depending on Artificial Intelligence (AI) systems to build various investment strategies. Support Vector Machine algorithm (SVM) is an advanced technique for classification, regression, and forecasting purposes which introduced by Cortes and Vapnik (1995). This study applies Support Vector Machine to predict the direction of GCC stock indexes movement using the historical prices. Data sample in this study covers the period from 2010 till 2013. This study compares Support Vector Machine (SVM) with other classification methods such as Logistic Regression and Random Forest. This study suggests that Support Vector Machine (SVM) perform well in predicting the direction of indexes movement in GCC stock markets where accuracy average rates are between 72.06% and 83.42% in all markets. In addition to that, Support Vector Machine outperform Random Forest algorithm in predicting the direction of GCC stock indexes. However, Logistic Regression outperforms Support Vector Machine and Random Forest in all GCC markets. The findings of this study suggest that applying SVM and other Artificial Intelligence techniques in the trading systems might attract more investors and bring more commission to the brokerage companies in GCC stock markets.
Description: DISSERTATION WITH DISTINCTION
URI: http://bspace.buid.ac.ae/handle/1234/631
Appears in Collections:Dissertations for Finance and Banking (FB)

Files in This Item:
File Description SizeFormat 
110001.pdf1.57 MBAdobe PDFThumbnail
View/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.