HA180 : Stock Market Index Forecasting using Wavelet Neural Network
Thesis > Central Library of Shahrood University > Industrial Engineering & Management > MSc > 2017
Authors:
Mozhgan Ameri [Author], Sayyed Mojtaba Mirlohi[Supervisor], Alireza Ahmadifard[Advisor]
Abstarct: In this study, we tried to forecast Tehran Exchange Price Index (TEPIX) for 100 steps ahead using Wavelet transforms and Artificial Neural Network. The daily closing price data with the Tehran Exchange Price Index from August 23, 2003 to August 23, 2015 are used to train the ANNs.1 to improve the effectiveness of prediction algorithms, we have trained two kinds of ANNs. The first one, uses TEPIX daily values for training and forecasting, the second one uses TEPIX daily values and 10 more variables like the daily prices of gold, oil, USD,GBP etc. .2 we have trained these two ANNs with different values of hidden neurons and dilations to find out which one has the best performance.3 we have forecasted 100 day ahead of TEPIX.
Keywords:
#TEPIX #Wavelet #Neural Networks #db5 Link
Keeping place: Central Library of Shahrood University
Visitor: