Mining for Profitable Patterns in the Stock Market
Yihua Philip Sheng (Southern Illinois University, USA), Wen-Chi Hou (Southern Illinois University, USA) and Zhong Chen (Shanghai JiaoTong University, PR China)
Copyright: © 2005
The stock market, like other economic phenomena, is a very complex system. Many factors, such as company news, interest rates, macro economic data, and investors’ hopes and fears, all affect its behavior (Pring, 1991; Sharpe, Alexander, & Bailey, 1999). Investors have longed for tools and algorithms to analyze and predict stock market movement. In this study, we combine a financial theory, the market efficiency theory, and a data mining technique to explore profitable trading patterns in the stock market. To observe the price oscillation of several consecutive trading days, we examine the K-lines, each of which represents a stock’s one-day movement. We will use a data mining technique with a heuristic rating algorithm to mine for reliable patterns indicating price rise or fall in the near future.