Please use this identifier to cite or link to this item:
Title: Expert systems for forecasting stock market indices
Keywords: Financial Forecasting; Financial Time Series; Stock Market Prediction; Stock Market Index; Feed-forward Neural Networks; Quasi-Newton Method
Issue Date: 14-Jan-2005
Citation: KOH CHUNG HAUR (2005-01-14). Expert systems for forecasting stock market indices. ScholarBank@NUS Repository.
Abstract: Financial time series forecasting continues to draw considerable attention from both within the academic community and the financial market practitioners. A forecaster with high accuracy will be of great help to both individual and institutional investors.Neural networks, having the capabilities in classifying and recognizing non-stationary and non-linear time series data, has become one of the most attractive tools for both forecasting researchers and practitioners.This thesis proposes component-based neural network model for forecasting financial indices and applies the model to perform one-day ahead predictions of stock market indices in New York, London, Hong Kong, Frankfurt and Singapore. Unlike other financial forecasting models, the proposed model directly uses the prices and returns of component stocks of the index as inputs for the prediction. For the neural network training, parallel quasi-Newton method is applied in order to speed up the training process. Experiments were conducted to determine the optimal training iteration, the optimal data set size, the optimal number of input neurons and the optimal number of hidden neurons. Computational results show that our proposed component-based neural network model obtained satisfactory accuracy for different stock market indices. The result has encouraged us to explore the similarities among these markets and the extensibility to use component-based neural network model for financial market indices prediction.
Appears in Collections:Master's Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
thesis.pdf655.44 kBAdobe PDF



Page view(s)

checked on Feb 10, 2019


checked on Feb 10, 2019

Google ScholarTM


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