Click for new scientific resources and news about Corona[COVID-19]

Paper Information

Journal:   IRANIAN JOURNAL OF APPLIED ECONOMICS   fall 2017 , Volume 7 , Number 22 #r00450; Page(s) 59 To 70.

Stock Price Trend Prediction via Technical Analysis Indicators, Using the Hybrid Method of Genetic Algorithm and Artificial Neural Network: Case Study Iran Khodro Stocks

Author(s):  Azarian Zeinab, Homayouni Seyed Mahdi*
* Department of Industrial Engineering, Lenjan Azad University, Isfahan, Iran
An accurate prediction of stock market trends is important for investors' financial decisions. Using a set of technical analysis indicators is one of the most widely used financial prediction methods. On one hand, determining the appropriate parameters for these indicators as well as their combination is a challenging decision for researchers. On the other hand, the non-linear and dynamic nature of the stock market trend lead to widespread use of nonlinear prediction methods such as artificial neural networks. Although technical analysis indicators have been used as an input of artificial neural networks, so far, the use of optimized parameters of technical analysis indicators as an artificial neural network input has not been studied. Since each stock has its own trend, using a set of default or identical parameters for all types of stocks is not accurate. In this research, the parameters of a set of technical analysis indicators are optimized using a genetic algorithm for a particular stock as input to the artificial neural network. This hybrid method is used to predict next day stock price trends. It is assumed that according to the forecast of next day trend, the investor decides to buy, sell, or hold the stocks. To evaluate the performance of the proposed hybrid method, an artificial neural network using the technical analysis indicators with the default parameters is also used to predict the stock price trend. These two methods are implemented for the actual data of Iran Khodro Company stock. Results show the superior performance of the hybrid method by 25. 1% decrease in the error of prediction.
Keyword(s): Prediction,Technical Analysis Indicators,Genetic Algorithm,Artificial Neural Network,Stock Exchange Market
  • ندارد
  Persian Abstract Yearly Visit 114
Latest on Blog
Enter SID Blog