Author Zhang, Ruying
Title Using scientific methods in data mining application.
Institute Thesis (M.Sc.) National Institute of Education, Nanyang Technological University
Year 2000
Supervisor Khoo, G. S. ; Xu, Shu Yan
Call no. QA276.4 Zha
Data mining is becoming more and more important in financial field.  The overview of data mining methods, its techniques and examples are discussed in this thesis.  The methodology of the most important data mining methods-Neural Networks (NNs) is focused on, including the basic concept of NNs, the mathematics of NNs and training of NNs etc.  The methodology and results of forecasting of slope and the mean of return using the back-propagation (BP) neural network are also described.  Based on the high accuracy forecasted results that we have obtained, we can conclude that using the neural network to forecast financial parameters is reasonable enough.  Simulations are done using forecasted results, and the  good results that are better than the market performance are achieved.