To search, Click below search items.

 

All Published Papers Search Service

Title

Neural Network Based Stock Market Forecasting

Author

Ahmed Ismail El-Hammady, Mohamed Aborizka

Citation

Vol. 11  No. 8  pp. 204-207

Abstract

We develop a useful prediction system in forecasting stock price for Egypt stock market (Egypt stock exchange weighted stock index, Commercial International Bank as CIB). The system is based on a recurrent neural network trained by using ARIMA analyses by differencing the raw data of the CIB series and then examining the ACF and PACF plots. The series can be identified as a nonlinear version of ARIMA (2, 1,2). Neural networks trained by using first differentiation of data. The networks trained by using history data for 6-years ago to predict 12 weeks market trend.

Keywords

Neural Network, Time Series, Forecast, ARIMA, Egypt Stock Market

URL

http://paper.ijcsns.org/07_book/201108/20110830.pdf