Techniques in financial time series prediction

• Main Author: Wu, Qingwei.
• Other Authors: Yap Kim Hui
• Format: Final Year Project
• Language: English
• Published: 2011
• Subjects: DRNTU::Social sciences::Economic theory
• Online Access: http://hdl.handle.net/10356/45864
• Institution: Nanyang Technological University

Stock markets around the world are affected by many highly correlated economic, political and even psychological factors. The interaction of these factors is in a very complex manner. The stock price time series usually contains the characteristics of high-noise and non-stationary, which makes classical statistical methods of price prediction incompetent. It is thus necessary to adopt more advanced forecasting techniques. Neural network is a type of nonlinear model that has proved to be effective for time series prediction. Therefore, in this study, multi-layer feedforward artificial neural network based on back-propagation algorithm is proposed and MATLAB is used to implement the design and training of the neural network. The model is trained with two years of historical data from January 2009 to December 2010 in order to predict the stock prices of General Electric. The accuracy of the forecasting model is studied by calculating the difference between raw data and simulation model results. The performance of the prediction model is evaluated using two widely used statistical metrics.